1#if 0
2CHANGES - changes for libpng
3
4Version 0.2
5  added reader into png.h
6  fixed small problems in stub file
7
8Version 0.3
9  added pull reader
10  split up pngwrite.c to several files
11  added pnglib.txt
12  added example.c
13  cleaned up writer, adding a few new transformations
14  fixed some bugs in writer
15  interfaced with zlib 0.5
16  added K&R support
17  added check for 64 KB blocks for 16-bit machines
18
19Version 0.4
20  cleaned up code and commented code
21  simplified time handling into png_time
22  created png_color_16 and png_color_8 to handle color needs
23  cleaned up color type defines
24  fixed various bugs
25  made various names more consistent
26  interfaced with zlib 0.71
27  cleaned up zTXt reader and writer (using zlib's Reset functions)
28  split transformations into pngrtran.c and pngwtran.c
29
30Version 0.5
31  interfaced with zlib 0.8
32  fixed many reading and writing bugs
33  saved using 3 spaces instead of tabs
34
35Version 0.6
36  added png_large_malloc() and png_large_free()
37  added png_size_t
38  cleaned up some compiler warnings
39  added png_start_read_image()
40
41Version 0.7
42  cleaned up lots of bugs
43  finished dithering and other stuff
44  added test program
45  changed name from pnglib to libpng
46
47Version 0.71 [June, 1995]
48  changed pngtest.png for zlib 0.93
49  fixed error in libpng.txt and example.c
50
51Version 0.8
52  cleaned up some bugs
53  added png_set_filler()
54  split up pngstub.c into pngmem.c, pngio.c, and pngerror.c
55  added #define's to remove unwanted code
56  moved png_info_init() to png.c
57  added old_size into png_realloc()
58  added functions to manually set filtering and compression info
59  changed compression parameters based on image type
60  optimized filter selection code
61  added version info
62  changed external functions passing floats to doubles (k&r problems?)
63  put all the configurable stuff in pngconf.h
64  enabled png_set_shift to work with paletted images on read
65  added png_read_update_info() - updates info structure with transformations
66
67Version 0.81 [August, 1995]
68  incorporated Tim Wegner's medium model code (thanks, Tim)
69
70Version 0.82 [September, 1995]
71  [unspecified changes]
72
73Version 0.85 [December, 1995]
74  added more medium model code (almost everything's a far)
75  added i/o, error, and memory callback functions
76  fixed some bugs (16-bit, 4-bit interlaced, etc.)
77  added first run progressive reader (barely tested)
78
79Version 0.86 [January, 1996]
80  fixed bugs
81  improved documentation
82
83Version 0.87 [January, 1996]
84  fixed medium model bugs
85  fixed other bugs introduced in 0.85 and 0.86
86  added some minor documentation
87
88Version 0.88 [January, 1996]
89  fixed progressive bugs
90  replaced tabs with spaces
91  cleaned up documentation
92  added callbacks for read/write and warning/error functions
93
94Version 0.89 [July, 1996]
95  Added new initialization API to make libpng work better with shared libs
96    we now have png_create_read_struct(), png_create_write_struct(),
97    png_create_info_struct(), png_destroy_read_struct(), and
98    png_destroy_write_struct() instead of the separate calls to
99    malloc and png_read_init(), png_info_init(), and png_write_init()
100  Changed warning/error callback functions to fix bug - this means you
101    should use the new initialization API if you were using the old
102    png_set_message_fn() calls, and that the old API no longer exists
103    so that people are aware that they need to change their code
104  Changed filter selection API to allow selection of multiple filters
105    since it didn't work in previous versions of libpng anyways
106  Optimized filter selection code
107  Fixed png_set_background() to allow using an arbitrary RGB color for
108    paletted images
109  Fixed gamma and background correction for paletted images, so
110    png_correct_palette is not needed unless you are correcting an
111    external palette (you will need to #define PNG_CORRECT_PALETTE_SUPPORTED
112    in pngconf.h) - if nobody uses this, it may disappear in the future.
113  Fixed bug with Borland 64K memory allocation (Alexander Lehmann)
114  Fixed bug in interlace handling (Smarasderagd, I think)
115  Added more error checking for writing and image to reduce invalid files
116  Separated read and write functions so that they won't both be linked
117    into a binary when only reading or writing functionality is used
118  New pngtest image also has interlacing and zTXt
119  Updated documentation to reflect new API
120
121Version 0.90 [January, 1997]
122  Made CRC errors/warnings on critical and ancillary chunks configurable
123  libpng will use the zlib CRC routines by (compile-time) default
124  Changed DOS small/medium model memory support - needs zlib 1.04 (Tim Wegner)
125  Added external C++ wrapper statements to png.h (Gilles Dauphin)
126  Allow PNG file to be read when some or all of file signature has already
127    been read from the beginning of the stream.  ****This affects the size
128    of info_struct and invalidates all programs that use a shared libpng****
129  Fixed png_filler() declarations
130  Fixed? background color conversions
131  Fixed order of error function pointers to match documentation
132  Current chunk name is now available in png_struct to reduce the number
133    of nearly identical error messages (will simplify multi-lingual
134    support when available)
135  Try to get ready for unknown-chunk callback functions:
136    - previously read critical chunks are flagged, so the chunk handling
137      routines can determine if the chunk is in the right place
138    - all chunk handling routines have the same prototypes, so we will
139      be able to handle all chunks via a callback mechanism
140  Try to fix Linux "setjmp" buffer size problems
141  Removed png_large_malloc, png_large_free, and png_realloc functions.
142
143Version 0.95 [March, 1997]
144  Fixed bug in pngwutil.c allocating "up_row" twice and "avg_row" never
145  Fixed bug in PNG file signature compares when start != 0
146  Changed parameter type of png_set_filler(...filler...) from png_byte
147    to png_uint_32
148  Added test for MACOS to ensure that both math.h and fp.h are not #included
149  Added macros for libpng to be compiled as a Windows DLL (Andreas Kupries)
150  Added "packswap" transformation, which changes the endianness of
151    packed-pixel bytes (Kevin Bracey)
152  Added "strip_alpha" transformation, which removes the alpha channel of
153    input images without using it (not necessarily a good idea)
154  Added "swap_alpha" transformation, which puts the alpha channel in front
155    of the color bytes instead of after
156  Removed all implicit variable tests which assume NULL == 0 (I think)
157  Changed several variables to "png_size_t" to show 16/32-bit limitations
158  Added new pCAL chunk read/write support
159  Added experimental filter selection weighting (Greg Roelofs)
160  Removed old png_set_rgbx() and png_set_xrgb() functions that have been
161     obsolete for about 2 years now (use png_set_filler() instead)
162  Added macros to read 16- and 32-bit ints directly from buffer, to be
163    used only on those systems that support it (namely PowerPC and 680x0)
164    With some testing, this may become the default for MACOS/PPC systems.
165  Only calculate CRC on data if we are going to use it
166  Added macros for zTXt compression type PNG_zTXt_COMPRESSION_???
167  Added macros for simple libpng debugging output selectable at compile time
168  Removed PNG_READ_END_MODE in progressive reader (Smarasderagd)
169  More description of info_struct in libpng.txt and png.h
170  More instructions in example.c
171  More chunk types tested in pngtest.c
172  Renamed pngrcb.c to pngset.c, and all png_read_<chunk> functions to be
173    png_set_<chunk>.  We now have corresponding png_get_<chunk>
174    functions in pngget.c to get information in info_ptr.  This isolates
175    the application from the internal organization of png_info_struct
176    (good for shared library implementations).
177
178Version 0.96 [May, 1997]
179  Fixed serious bug with < 8bpp images introduced in 0.95
180  Fixed 256-color transparency bug (Greg Roelofs)
181  Fixed up documentation (Greg Roelofs, Laszlo Nyul)
182  Fixed "error" in pngconf.h for Linux setjmp() behavior
183  Fixed DOS medium model support (Tim Wegner)
184  Fixed png_check_keyword() for case with error in static string text
185  Added read of CRC after IEND chunk for embedded PNGs (Laszlo Nyul)
186  Added typecasts to quiet compiler errors
187  Added more debugging info
188
189Version 0.97 [January, 1998]
190  Removed PNG_USE_OWN_CRC capability
191  Relocated png_set_crc_action from pngrutil.c to pngrtran.c
192  Fixed typecasts of "new_key", etc. (Andreas Dilger)
193  Added RFC 1152 [sic] date support
194  Fixed bug in gamma handling of 4-bit grayscale
195  Added 2-bit grayscale gamma handling (Glenn R-P)
196  Added more typecasts. 65536L becomes (png_uint_32)65536L, etc. (Glenn R-P)
197  Minor corrections in libpng.txt
198  Added simple sRGB support (Glenn R-P)
199  Easier conditional compiling, e.g.,
200    define PNG_READ/WRITE_NOT_FULLY_SUPPORTED;
201    all configurable options can be selected from command-line instead
202    of having to edit pngconf.h (Glenn R-P)
203  Fixed memory leak in pngwrite.c (free info_ptr->text) (Glenn R-P)
204  Added more conditions for png_do_background, to avoid changing
205    black pixels to background when a background is supplied and
206    no pixels are transparent
207  Repaired PNG_NO_STDIO behavior
208  Tested NODIV support and made it default behavior (Greg Roelofs)
209  Added "-m" option and PNGTEST_DEBUG_MEMORY to pngtest (John Bowler)
210  Regularized version numbering scheme and bumped shared-library major
211    version number to 2 to avoid problems with libpng 0.89 apps
212    (Greg Roelofs)
213
214Version 0.98 [January, 1998]
215  Cleaned up some typos in libpng.txt and in code documentation
216  Fixed memory leaks in pCAL chunk processing (Glenn R-P and John Bowler)
217  Cosmetic change "display_gamma" to "screen_gamma" in pngrtran.c
218  Changed recommendation about file_gamma for PC images to .51 from .45,
219    in example.c and libpng.txt, added comments to distinguish between
220    screen_gamma, viewing_gamma, and display_gamma.
221  Changed all references to RFC1152 to read RFC1123 and changed the
222    PNG_TIME_RFC1152_SUPPORTED macro to PNG_TIME_RFC1123_SUPPORTED
223  Added png_invert_alpha capability (Glenn R-P -- suggestion by Jon Vincent)
224  Changed srgb_intent from png_byte to int to avoid compiler bugs
225
226Version 0.99 [January 30, 1998]
227  Free info_ptr->text instead of end_info_ptr->text in pngread.c (John Bowler)
228  Fixed a longstanding "packswap" bug in pngtrans.c
229  Fixed some inconsistencies in pngconf.h that prevented compiling with
230    PNG_READ_GAMMA_SUPPORTED and PNG_READ_hIST_SUPPORTED undefined
231  Fixed some typos and made other minor rearrangement of libpng.txt (Andreas)
232  Changed recommendation about file_gamma for PC images to .50 from .51 in
233    example.c and libpng.txt, and changed file_gamma for sRGB images to .45
234  Added a number of functions to access information from the png structure
235    png_get_image_height(), etc. (Glenn R-P, suggestion by Brad Pettit)
236  Added TARGET_MACOS similar to zlib-1.0.8
237  Define PNG_ALWAYS_EXTERN when __MWERKS__ && WIN32 are defined
238  Added type casting to all png_malloc() function calls
239
240Version 0.99a [January 31, 1998]
241  Added type casts and parentheses to all returns that return a value.(Tim W.)
242
243Version 0.99b [February 4, 1998]
244  Added type cast png_uint_32 on malloc function calls where needed.
245  Changed type of num_hist from png_uint_32 to int (same as num_palette).
246  Added checks for rowbytes overflow, in case png_size_t is less than 32 bits.
247  Renamed makefile.elf to makefile.lnx.
248
249Version 0.99c [February 7, 1998]
250  More type casting.  Removed erroneous overflow test in pngmem.c.
251  Added png_buffered_memcpy() and png_buffered_memset(), apply them to rowbytes.
252  Added UNIX manual pages libpng.3 (incorporating libpng.txt) and  png.5.
253
254Version 0.99d [February 11, 1998]
255  Renamed "far_to_near()" "png_far_to_near()"
256  Revised libpng.3
257  Version 99c "buffered" operations didn't work as intended.  Replaced them
258    with png_memcpy_check() and png_memset_check().
259  Added many "if (png_ptr == NULL) return" to quell compiler warnings about
260    unused png_ptr, mostly in pngget.c and pngset.c.
261  Check for overlength tRNS chunk present when indexed-color PLTE is read.
262  Cleaned up spelling errors in libpng.3/libpng.txt
263  Corrected a problem with png_get_tRNS() which returned undefined trans array
264
265Version 0.99e [February 28, 1998]
266  Corrected png_get_tRNS() again.
267  Add parentheses for easier reading of pngget.c, fixed "||" should be "&&".
268  Touched up example.c to make more of it compileable, although the entire
269    file still can't be compiled (Willem van Schaik)
270  Fixed a bug in png_do_shift() (Bryan Tsai)
271  Added a space in png.h prototype for png_write_chunk_start()
272  Replaced pngtest.png with one created with zlib 1.1.1
273  Changed pngtest to report PASS even when file size is different (Jean-loup G.)
274  Corrected some logic errors in png_do_invert_alpha() (Chris Patterson)
275
276Version 0.99f [March 5, 1998]
277  Corrected a bug in pngpread() introduced in version 99c (Kevin Bracey)
278  Moved makefiles into a "scripts" directory, and added INSTALL instruction file
279  Added makefile.os2 and pngos2.def (A. Zabolotny) and makefile.s2x (W. Sebok)
280  Added pointers to "note on libpng versions" in makefile.lnx and README
281  Added row callback feature when reading and writing nonprogressive rows
282    and added a test of this feature in pngtest.c
283  Added user transform callbacks, with test of the feature in pngtest.c
284
285Version 0.99g [March 6, 1998, morning]
286  Minor changes to pngtest.c to suppress compiler warnings.
287  Removed "beta" language from documentation.
288
289Version 0.99h [March 6, 1998, evening]
290  Minor changes to previous minor changes to pngtest.c
291  Changed PNG_READ_NOT_FULLY_SUPPORTED to PNG_READ_TRANSFORMS_NOT_SUPPORTED
292    and added PNG_PROGRESSIVE_READ_NOT_SUPPORTED macro
293  Added user transform capability
294
295Version 1.00 [March 7, 1998]
296  Changed several typedefs in pngrutil.c
297  Added makefile.wat (Pawel Mrochen), updated makefile.tc3 (Willem van Schaik)
298  Replaced "while(1)" with "for(;;)"
299  Added PNGARG() to prototypes in pngtest.c and removed some prototypes
300  Updated some of the makefiles (Tom Lane)
301  Changed some typedefs (s_start, etc.) in pngrutil.c
302  Fixed dimensions of "short_months" array in pngwrite.c
303  Replaced ansi2knr.c with the one from jpeg-v6
304
305Version 1.0.0 [March 8, 1998]
306  Changed name from 1.00 to 1.0.0 (Adam Costello)
307  Added smakefile.ppc (with SCOPTIONS.ppc) for Amiga PPC (Andreas Kleinert)
308
309Version 1.0.0a [March 9, 1998]
310  Fixed three bugs in pngrtran.c to make gamma+background handling consistent
311    (Greg Roelofs)
312  Changed format of the PNG_LIBPNG_VER integer to xyyzz instead of xyz
313    for major, minor, and bugfix releases.  This is 10001. (Adam Costello,
314    Tom Lane)
315  Make months range from 1-12 in png_convert_to_rfc1123
316
317Version 1.0.0b [March 13, 1998]
318  Quieted compiler complaints about two empty "for" loops in pngrutil.c
319  Minor changes to makefile.s2x
320  Removed #ifdef/#endif around a png_free() in pngread.c
321
322Version 1.0.1 [March 14, 1998]
323  Changed makefile.s2x to reduce security risk of using a relative pathname
324  Fixed some typos in the documentation (Greg).
325  Fixed a problem with value of "channels" returned by png_read_update_info()
326
327Version 1.0.1a [April 21, 1998]
328  Optimized Paeth calculations by replacing abs() function calls with intrinsics
329  plus other loop optimizations. Improves avg decoding speed by about 20%.
330  Commented out i386istic "align" compiler flags in makefile.lnx.
331  Reduced the default warning level in some makefiles, to make them consistent.
332  Removed references to IJG and JPEG in the ansi2knr.c copyright statement.
333  Fixed a bug in png_do_strip_filler with XXRRGGBB => RRGGBB transformation.
334  Added grayscale and 16-bit capability to png_do_read_filler().
335  Fixed a bug in pngset.c, introduced in version 0.99c, that sets rowbytes
336    too large when writing an image with bit_depth < 8 (Bob Dellaca).
337  Corrected some bugs in the experimental weighted filtering heuristics.
338  Moved a misplaced pngrutil code block that truncates tRNS if it has more
339    than num_palette entries -- test was done before num_palette was defined.
340  Fixed a png_convert_to_rfc1123() bug that converts day 31 to 0 (Steve Eddins).
341  Changed compiler flags in makefile.wat for better optimization
342    (Pawel Mrochen).
343
344Version 1.0.1b [May 2, 1998]
345  Relocated png_do_gray_to_rgb() within png_do_read_transformations() (Greg).
346  Relocated the png_composite macros from pngrtran.c to png.h (Greg).
347  Added makefile.sco (contributed by Mike Hopkirk).
348  Fixed two bugs (missing definitions of "istop") introduced in libpng-1.0.1a.
349  Fixed a bug in pngrtran.c that would set channels=5 under some circumstances.
350  More work on the Paeth-filtering, achieving imperceptible speedup
351    (A Kleinert).
352  More work on loop optimization which may help when compiled with C++
353    compilers.
354  Added warnings when people try to use transforms they've defined out.
355  Collapsed 4 "i" and "c" loops into single "i" loops in pngrtran and pngwtran.
356  Revised paragraph about png_set_expand() in libpng.txt and libpng.3 (Greg)
357
358Version 1.0.1c [May 11, 1998]
359  Fixed a bug in pngrtran.c (introduced in libpng-1.0.1a) where the masks for
360    filler bytes should have been 0xff instead of 0xf.
361  Added max_pixel_depth=32 in pngrutil.c when using FILLER with palette images.
362  Moved PNG_WRITE_WEIGHTED_FILTER_SUPPORTED and PNG_WRITE_FLUSH_SUPPORTED
363    out of the PNG_WRITE_TRANSFORMS_NOT_SUPPORTED block of pngconf.h
364  Added "PNG_NO_WRITE_TRANSFORMS" etc., as alternatives for *_NOT_SUPPORTED,
365    for consistency, in pngconf.h
366  Added individual "ifndef PNG_NO_[CAPABILITY]" in pngconf.h to make it easier
367    to remove unwanted capabilities via the compile line
368  Made some corrections to grammar (which, it's) in documentation (Greg).
369  Corrected example.c, use of row_pointers in png_write_image().
370
371Version 1.0.1d [May 24, 1998]
372  Corrected several statements that used side effects illegally in pngrutil.c
373    and pngtrans.c, that were introduced in version 1.0.1b
374  Revised png_read_rows() to avoid repeated if-testing for NULL (A Kleinert)
375  More corrections to example.c, use of row_pointers in png_write_image()
376    and png_read_rows().
377  Added pngdll.mak and pngdef.pas to scripts directory, contributed by
378    Bob Dellaca, to make a png32bd.dll with Borland C++ 4.5
379  Fixed error in example.c with png_set_text: num_text is 3, not 2 (Guido V.)
380  Changed several loops from count-down to count-up, for consistency.
381
382Version 1.0.1e [June 6, 1998]
383  Revised libpng.txt and libpng.3 description of png_set_read|write_fn(), and
384    added warnings when people try to set png_read_fn and png_write_fn in
385    the same structure.
386  Added a test such that png_do_gamma will be done when num_trans==0
387    for truecolor images that have defined a background.  This corrects an
388    error that was introduced in libpng-0.90 that can cause gamma processing
389    to be skipped.
390  Added tests in png.h to include "trans" and "trans_values" in structures
391    when PNG_READ_BACKGROUND_SUPPORTED or PNG_READ_EXPAND_SUPPORTED is defined.
392  Add png_free(png_ptr->time_buffer) in png_destroy_read_struct()
393  Moved png_convert_to_rfc_1123() from pngwrite.c to png.c
394  Added capability for user-provided malloc_fn() and free_fn() functions,
395    and revised pngtest.c to demonstrate their use, replacing the
396    PNGTEST_DEBUG_MEM feature.
397  Added makefile.w32, for Microsoft C++ 4.0 and later (Tim Wegner).
398
399Version 1.0.2 [June 14, 1998]
400  Fixed two bugs in makefile.bor .
401
402Version 1.0.2a [December 30, 1998]
403  Replaced and extended code that was removed from png_set_filler() in 1.0.1a.
404  Fixed a bug in png_do_filler() that made it fail to write filler bytes in
405    the left-most pixel of each row (Kevin Bracey).
406  Changed "static pngcharp tIME_string" to "static char tIME_string[30]"
407    in pngtest.c (Duncan Simpson).
408  Fixed a bug in pngtest.c that caused pngtest to try to write a tIME chunk
409    even when no tIME chunk was present in the source file.
410  Fixed a problem in pngrutil.c: gray_to_rgb didn't always work with 16-bit.
411  Fixed a problem in png_read_push_finish_row(), which would not skip some
412    passes that it should skip, for images that are less than 3 pixels high.
413  Interchanged the order of calls to png_do_swap() and png_do_shift()
414    in pngwtran.c (John Cromer).
415  Added #ifdef PNG_DEBUG/#endif surrounding use of PNG_DEBUG in png.h .
416  Changed "bad adaptive filter type" from error to warning in pngrutil.c .
417  Fixed a documentation error about default filtering with 8-bit indexed-color.
418  Separated the PNG_NO_STDIO macro into PNG_NO_STDIO and PNG_NO_CONSOLE_IO
419    (L. Peter Deutsch).
420  Added png_set_rgb_to_gray() and png_get_rgb_to_gray_status() functions.
421  Added png_get_copyright() and png_get_header_version() functions.
422  Revised comments on png_set_progressive_read_fn() in libpng.txt and example.c
423  Added information about debugging in libpng.txt and libpng.3 .
424  Changed "ln -sf" to "ln -s -f" in makefile.s2x, makefile.lnx, and
425    makefile.sco.
426  Removed lines after Dynamic Dependencies" in makefile.aco .
427  Revised makefile.dec to make a shared library (Jeremie Petit).
428  Removed trailing blanks from all files.
429
430Version 1.0.2a [January 6, 1999]
431  Removed misplaced #endif and #ifdef PNG_NO_EXTERN near the end of png.h
432  Added "if" tests to silence complaints about unused png_ptr in png.h and png.c
433  Changed "check_if_png" function in example.c to return true (nonzero) if PNG.
434  Changed libpng.txt to demonstrate png_sig_cmp() instead of png_check_sig()
435    which is obsolete.
436
437Version 1.0.3 [January 14, 1999]
438  Added makefile.hux, for Hewlett Packard HPUX 10.20 and 11.00 (Jim Rice)
439  Added a statement of Y2K compliance in png.h, libpng.3, and Y2KINFO.
440
441Version 1.0.3a [August 12, 1999]
442  Added check for PNG_READ_INTERLACE_SUPPORTED in pngread.c; issue a warning
443     if an attempt is made to read an interlaced image when it's not supported.
444  Added check if png_ptr->trans is defined before freeing it in pngread.c
445  Modified the Y2K statement to include versions back to version 0.71
446  Fixed a bug in the check for valid IHDR bit_depth/color_types in pngrutil.c
447  Modified makefile.wat (added -zp8 flag, ".symbolic", changed some comments)
448  Replaced leading blanks with tab characters in makefile.hux
449  Changed "dworkin.wustl.edu" to "ccrc.wustl.edu" in various documents.
450  Changed (float)red and (float)green to (double)red, (double)green
451     in png_set_rgb_to_gray() to avoid "promotion" problems in AIX.
452  Fixed a bug in pngconf.h that omitted <stdio.h> when PNG_DEBUG==0 (K Bracey).
453  Reformatted libpng.3 and libpngpf.3 with proper fonts (script by J. vanZandt).
454  Updated documentation to refer to the PNG-1.2 specification.
455  Removed ansi2knr.c and left pointers to the latest source for ansi2knr.c
456    in makefile.knr, INSTALL, and README (L. Peter Deutsch)
457  Fixed bugs in calculation of the length of rowbytes when adding alpha
458    channels to 16-bit images, in pngrtran.c (Chris Nokleberg)
459  Added function png_set_user_transform_info() to store user_transform_ptr,
460    user_depth, and user_channels into the png_struct, and a function
461    png_get_user_transform_ptr() to retrieve the pointer (Chris Nokleberg)
462  Added function png_set_empty_plte_permitted() to make libpng useable
463    in MNG applications.
464  Corrected the typedef for png_free_ptr in png.h (Jesse Jones).
465  Correct gamma with srgb is 45455 instead of 45000 in pngrutil.c, to be
466    consistent with PNG-1.2, and allow variance of 500 before complaining.
467  Added assembler code contributed by Intel in file pngvcrd.c and modified
468    makefile.w32 to use it (Nirav Chhatrapati, INTEL Corporation,
469    Gilles Vollant)
470  Changed "ln -s -f" to "ln -f -s" in the makefiles to make Solaris happy.
471  Added some aliases for png_set_expand() in pngrtran.c, namely
472    png_set_expand_PLTE(), png_set_expand_depth(), and png_set_expand_tRNS()
473    (Greg Roelofs, in "PNG: The Definitive Guide").
474  Added makefile.beo for BEOS on X86, contributed by Sander Stok.
475
476Version 1.0.3b [August 26, 1999]
477  Replaced 2147483647L several places with PNG_MAX_UINT macro, defined in png.h
478  Changed leading blanks to tabs in all makefiles.
479  Define PNG_USE_PNGVCRD in makefile.w32, to get MMX assembler code.
480  Made alternate versions of  png_set_expand() in pngrtran.c, namely
481    png_set_gray_1_2_4_to_8, png_set_palette_to_rgb, and png_set_tRNS_to_alpha
482    (Greg Roelofs, in "PNG: The Definitive Guide").  Deleted the 1.0.3a aliases.
483  Relocated start of 'extern "C"' block in png.h so it doesn't include pngconf.h
484  Revised calculation of num_blocks in pngmem.c to avoid a potentially
485    negative shift distance, whose results are undefined in the C language.
486  Added a check in pngset.c to prevent writing multiple tIME chunks.
487  Added a check in pngwrite.c to detect invalid small window_bits sizes.
488
489Version 1.0.3d [September 4, 1999]
490  Fixed type casting of igamma in pngrutil.c
491  Added new png_expand functions to scripts/pngdef.pas and pngos2.def
492  Added a demo read_user_transform_fn that examines the row filters in pngtest.c
493
494Version 1.0.4 [September 24, 1999]
495  Define PNG_ALWAYS_EXTERN in pngconf.h if __STDC__ is defined
496  Delete #define PNG_INTERNAL and include "png.h" from pngasmrd.h
497  Made several minor corrections to pngtest.c
498  Renamed the makefiles with longer but more user friendly extensions.
499  Copied the PNG copyright and license to a separate LICENSE file.
500  Revised documentation, png.h, and example.c to remove reference to
501    "viewing_gamma" which no longer appears in the PNG specification.
502  Revised pngvcrd.c to use MMX code for interlacing only on the final pass.
503  Updated pngvcrd.c to use the faster C filter algorithms from libpng-1.0.1a
504  Split makefile.win32vc into two versions, makefile.vcawin32 (uses MMX
505    assembler code) and makefile.vcwin32 (doesn't).
506  Added a CPU timing report to pngtest.c (enabled by defining PNGTEST_TIMING)
507  Added a copy of pngnow.png to the distribution.
508
509Version 1.0.4a [September 25, 1999]
510  Increase max_pixel_depth in pngrutil.c if a user transform needs it.
511  Changed several division operations to right-shifts in pngvcrd.c
512
513Version 1.0.4b [September 30, 1999]
514  Added parentheses in line 3732 of pngvcrd.c
515  Added a comment in makefile.linux warning about buggy -O3 in pgcc 2.95.1
516
517Version 1.0.4c [October 1, 1999]
518  Added a "png_check_version" function in png.c and pngtest.c that will generate
519    a helpful compiler error if an old png.h is found in the search path.
520  Changed type of png_user_transform_depth|channels from int to png_byte.
521
522Version 1.0.4d [October 6, 1999]
523  Changed 0.45 to 0.45455 in png_set_sRGB()
524  Removed unused PLTE entries from pngnow.png
525  Re-enabled some parts of pngvcrd.c (png_combine_row) that work properly.
526
527Version 1.0.4e [October 10, 1999]
528  Fixed sign error in pngvcrd.c (Greg Roelofs)
529  Replaced some instances of memcpy with simple assignments in pngvcrd (GR-P)
530
531Version 1.0.4f [October 15, 1999]
532  Surrounded example.c code with #if 0 .. #endif to prevent people from
533    inadvertently trying to compile it.
534  Changed png_get_header_version() from a function to a macro in png.h
535  Added type casting mostly in pngrtran.c and pngwtran.c
536  Removed some pointless "ptr = NULL" in pngmem.c
537  Added a "contrib" directory containing the source code from Greg's book.
538
539Version 1.0.5 [October 15, 1999]
540  Minor editing of the INSTALL and README files.
541
542Version 1.0.5a [October 23, 1999]
543  Added contrib/pngsuite and contrib/pngminus (Willem van Schaik)
544  Fixed a typo in the png_set_sRGB() function call in example.c (Jan Nijtmans)
545  Further optimization and bugfix of pngvcrd.c
546  Revised pngset.c so that it does not allocate or free memory in the user's
547    text_ptr structure.  Instead, it makes its own copy.
548  Created separate write_end_info_struct in pngtest.c for a more severe test.
549  Added code in pngwrite.c to free info_ptr->text[i].key to stop a memory leak.
550
551Version 1.0.5b [November 23, 1999]
552  Moved PNG_FLAG_HAVE_CHUNK_HEADER, PNG_FLAG_BACKGROUND_IS_GRAY and
553    PNG_FLAG_WROTE_tIME from flags to mode.
554  Added png_write_info_before_PLTE() function.
555  Fixed some typecasting in contrib/gregbook/*.c
556  Updated scripts/makevms.com and added makevms.com to contrib/gregbook
557    and contrib/pngminus (Martin Zinser)
558
559Version 1.0.5c [November 26, 1999]
560  Moved png_get_header_version from png.h to png.c, to accommodate ansi2knr.
561  Removed all global arrays (according to PNG_NO_GLOBAL_ARRAYS macro), to
562    accommodate making DLL's: Moved usr_png_ver from global variable to function
563    png_get_header_ver() in png.c.  Moved png_sig to png_sig_bytes in png.c and
564    eliminated use of png_sig in pngwutil.c.  Moved the various png_CHNK arrays
565    into pngtypes.h.  Eliminated use of global png_pass arrays.  Declared the
566    png_CHNK and png_pass arrays to be "const".  Made the global arrays
567    available to applications (although none are used in libpng itself) when
568    PNG_NO_GLOBAL_ARRAYS is not defined or when PNG_GLOBAL_ARRAYS is defined.
569  Removed some extraneous "-I" from contrib/pngminus/makefile.std
570  Changed the PNG_sRGB_INTENT macros in png.h to be consistent with PNG-1.2.
571  Change PNG_SRGB_INTENT to PNG_sRGB_INTENT in libpng.txt and libpng.3
572
573Version 1.0.5d [November 29, 1999]
574  Add type cast (png_const_charp) two places in png.c
575  Eliminated pngtypes.h; use macros instead to declare PNG_CHNK arrays.
576  Renamed "PNG_GLOBAL_ARRAYS" to "PNG_USE_GLOBAL_ARRAYS" and made available
577    to applications a macro "PNG_USE_LOCAL_ARRAYS".
578  comment out (with #ifdef) all the new declarations when
579    PNG_USE_GLOBAL_ARRAYS is defined.
580  Added PNG_EXPORT_VAR macro to accommodate making DLL's.
581
582Version 1.0.5e [November 30, 1999]
583  Added iCCP, iTXt, and sPLT support; added "lang" member to the png_text
584    structure; refactored the inflate/deflate support to make adding new chunks
585    with trailing compressed parts easier in the future, and added new functions
586    png_free_iCCP, png_free_pCAL, png_free_sPLT, png_free_text, png_get_iCCP,
587    png_get_spalettes, png_set_iCCP, png_set_spalettes (Eric S. Raymond).
588  NOTE: Applications that write text chunks MUST define png_text->lang
589    before calling png_set_text(). It must be set to NULL if you want to
590    write tEXt or zTXt chunks.  If you want your application to be able to
591    run with older versions of libpng, use
592
593      #ifdef PNG_iTXt_SUPPORTED
594         png_text[i].lang = NULL;
595      #endif
596
597  Changed png_get_oFFs() and png_set_oFFs() to use signed rather than unsigned
598    offsets (Eric S. Raymond).
599  Combined PNG_READ_cHNK_SUPPORTED and PNG_WRITE_cHNK_SUPPORTED macros into
600    PNG_cHNK_SUPPORTED and combined the three types of PNG_text_SUPPORTED
601    macros, leaving the separate macros also available.
602  Removed comments on #endifs at the end of many short, non-nested #if-blocks.
603
604Version 1.0.5f [December 6, 1999]
605  Changed makefile.solaris to issue a warning about potential problems when
606    the ucb "ld" is in the path ahead of the ccs "ld".
607  Removed "- [date]" from the "synopsis" line in libpng.3 and libpngpf.3.
608  Added sCAL chunk support (Eric S. Raymond).
609
610Version 1.0.5g [December 7, 1999]
611  Fixed "png_free_spallettes" typo in png.h
612  Added code to handle new chunks in pngpread.c
613  Moved PNG_CHNK string macro definitions outside of PNG_NO_EXTERN block
614  Added "translated_key" to png_text structure and png_write_iTXt().
615  Added code in pngwrite.c to work around a newly discovered zlib bug.
616
617Version 1.0.5h [December 10, 1999]
618  NOTE: regarding the note for version 1.0.5e, the following must also
619    be included in your code:
620        png_text[i].translated_key = NULL;
621  Unknown chunk handling is now supported.
622  Option to eliminate all floating point support was added.  Some new
623    fixed-point functions such as png_set_gAMA_fixed() were added.
624  Expanded tabs and removed trailing blanks in source files.
625
626Version 1.0.5i [December 13, 1999]
627  Added some type casts to silence compiler warnings.
628  Renamed "png_free_spalette" to "png_free_spalettes" for consistency.
629  Removed leading blanks from a #define in pngvcrd.c
630  Added some parameters to the new png_set_keep_unknown_chunks() function.
631  Added a test for up->location != 0 in the first instance of writing
632    unknown chunks in pngwrite.c
633  Changed "num" to "i" in png_free_spalettes() and png_free_unknowns() to
634    prevent recursion.
635  Added png_free_hIST() function.
636  Various patches to fix bugs in the sCAL and integer cHRM processing,
637    and to add some convenience macros for use with sCAL.
638
639Version 1.0.5j [December 21, 1999]
640  Changed "unit" parameter of png_write_sCAL from png_byte to int, to work
641    around buggy compilers.
642  Added new type "png_fixed_point" for integers that hold float*100000 values
643  Restored backward compatibility of tEXt/zTXt chunk processing:
644    Restored the first four members of png_text to the same order as v.1.0.5d.
645    Added members "lang_key" and "itxt_length" to png_text struct.  Set
646    text_length=0 when "text" contains iTXt data.  Use the "compression"
647    member to distinguish among tEXt/zTXt/iTXt types.  Added
648    PNG_ITXT_COMPRESSION_NONE (1) and PNG_ITXT_COMPRESSION_zTXt(2) macros.
649    The "Note" above, about backward incompatibility of libpng-1.0.5e, no
650    longer applies.
651  Fixed png_read|write_iTXt() to read|write parameters in the right order,
652    and to write the iTXt chunk after IDAT if it appears in the end_ptr.
653  Added pnggccrd.c, version of pngvcrd.c Intel assembler for gcc (Greg Roelofs)
654  Reversed the order of trying to write floating-point and fixed-point gAMA.
655
656Version 1.0.5k [December 27, 1999]
657  Added many parentheses, e.g., "if (a && b & c)" becomes "if (a && (b & c))"
658  Added png_handle_as_unknown() function (Glenn)
659  Added png_free_chunk_list() function and chunk_list and num_chunk_list members
660    of png_ptr.
661  Eliminated erroneous warnings about multiple sPLT chunks and sPLT-after-PLTE.
662  Fixed a libpng-1.0.5h bug in pngrutil.c that was issuing erroneous warnings
663    about ignoring incorrect gAMA with sRGB (gAMA was in fact not ignored)
664  Added png_free_tRNS(); png_set_tRNS() now malloc's its own trans array (ESR).
665  Define png_get_int_32 when oFFs chunk is supported as well as when pCAL is.
666  Changed type of proflen from png_int_32 to png_uint_32 in png_get_iCCP().
667
668Version 1.0.5l [January 1, 2000]
669  Added functions png_set_read_user_chunk_fn() and png_get_user_chunk_ptr()
670    for setting a callback function to handle unknown chunks and for
671    retrieving the associated user pointer (Glenn).
672
673Version 1.0.5m [January 7, 2000]
674  Added high-level functions png_read_png(), png_write_png(), png_free_pixels().
675
676Version 1.0.5n [January 9, 2000]
677  Added png_free_PLTE() function, and modified png_set_PLTE() to malloc its
678    own memory for info_ptr->palette.  This makes it safe for the calling
679    application to free its copy of the palette any time after it calls
680    png_set_PLTE().
681
682Version 1.0.5o [January 20, 2000]
683  Cosmetic changes only (removed some trailing blanks and TABs)
684
685Version 1.0.5p [January 31, 2000]
686  Renamed pngdll.mak to makefile.bd32
687  Cosmetic changes in pngtest.c
688
689Version 1.0.5q [February 5, 2000]
690  Relocated the makefile.solaris warning about PATH problems.
691  Fixed pngvcrd.c bug by pushing/popping registers in mmxsupport (Bruce Oberg)
692  Revised makefile.gcmmx
693  Added PNG_SETJMP_SUPPORTED, PNG_SETJMP_NOT_SUPPORTED, and PNG_ABORT() macros
694
695Version 1.0.5r [February 7, 2000]
696  Removed superfluous prototype for png_get_itxt from png.h
697  Fixed a bug in pngrtran.c that improperly expanded the background color.
698  Return *num_text=0 from png_get_text() when appropriate, and fix documentation
699    of png_get_text() in libpng.txt/libpng.3.
700
701Version 1.0.5s [February 18, 2000]
702  Added "png_jmp_env()" macro to pngconf.h, to help people migrate to the
703    new error handler that's planned for the next libpng release, and changed
704    example.c, pngtest.c, and contrib programs to use this macro.
705  Revised some of the DLL-export macros in pngconf.h (Greg Roelofs)
706  Fixed a bug in png_read_png() that caused it to fail to expand some images
707    that it should have expanded.
708  Fixed some mistakes in the unused and undocumented INCH_CONVERSIONS functions
709    in pngget.c
710  Changed the allocation of palette, history, and trans arrays back to
711    the version 1.0.5 method (linking instead of copying) which restores
712    backward compatibility with version 1.0.5.  Added some remarks about
713    that in example.c.  Added "free_me" member to info_ptr and png_ptr
714    and added png_free_data() function.
715  Updated makefile.linux and makefile.gccmmx to make directories conditionally.
716  Made cosmetic changes to pngasmrd.h
717  Added png_set_rows() and png_get_rows(), for use with png_read|write_png().
718  Modified png_read_png() to allocate info_ptr->row_pointers only if it
719    hasn't already been allocated.
720
721Version 1.0.5t [March 4, 2000]
722  Changed png_jmp_env() migration aiding macro to png_jmpbuf().
723  Fixed "interlace" typo (should be "interlaced") in contrib/gregbook/read2-x.c
724  Fixed bug with use of PNG_BEFORE_IHDR bit in png_ptr->mode, introduced when
725    PNG_FLAG_HAVE_CHUNK_HEADER was moved into png_ptr->mode in version 1.0.5b
726  Files in contrib/gregbook were revised to use png_jmpbuf() and to select
727    a 24-bit visual if one is available, and to allow abbreviated options.
728  Files in contrib/pngminus were revised to use the png_jmpbuf() macro.
729  Removed spaces in makefile.linux and makefile.gcmmx, introduced in 1.0.5s
730
731Version 1.0.5u [March 5, 2000]
732  Simplified the code that detects old png.h in png.c and pngtest.c
733  Renamed png_spalette (_p, _pp) to png_sPLT_t (_tp, _tpp)
734  Increased precision of rgb_to_gray calculations from 8 to 15 bits and
735    added png_set_rgb_to_gray_fixed() function.
736  Added makefile.bc32 (32-bit Borland C++, C mode)
737
738Version 1.0.5v [March 11, 2000]
739  Added some parentheses to the png_jmpbuf macro definition.
740  Updated references to the zlib home page, which has moved to freesoftware.com.
741  Corrected bugs in documentation regarding png_read_row() and png_write_row().
742  Updated documentation of png_rgb_to_gray calculations in libpng.3/libpng.txt.
743  Renamed makefile.borland,turboc3 back to makefile.bor,tc3 as in version 1.0.3,
744    revised borland makefiles; added makefile.ibmvac3 and makefile.gcc (Cosmin)
745
746Version 1.0.6 [March 20, 2000]
747  Minor revisions of makefile.bor, libpng.txt, and gregbook/rpng2-win.c
748  Added makefile.sggcc (SGI IRIX with gcc)
749
750Version 1.0.6d [April 7, 2000]
751  Changed sprintf() to strcpy() in png_write_sCAL_s() to work without STDIO
752  Added data_length parameter to png_decompress_chunk() function
753  Revised documentation to remove reference to abandoned png_free_chnk functions
754  Fixed an error in png_rgb_to_gray_fixed()
755  Revised example.c, usage of png_destroy_write_struct().
756  Renamed makefile.ibmvac3 to makefile.ibmc, added libpng.icc IBM project file
757  Added a check for info_ptr->free_me&PNG_FREE_TEXT when freeing text in png.c
758  Simplify png_sig_bytes() function to remove use of non-ISO-C strdup().
759
760Version 1.0.6e [April 9, 2000]
761  Added png_data_freer() function.
762  In the code that checks for over-length tRNS chunks, added check of
763    info_ptr->num_trans as well as png_ptr->num_trans (Matthias Benckmann)
764  Minor revisions of libpng.txt/libpng.3.
765  Check for existing data and free it if the free_me flag is set, in png_set_*()
766    and png_handle_*().
767  Only define PNG_WEIGHTED_FILTERS_SUPPORTED when PNG_FLOATING_POINT_SUPPORTED
768    is defined.
769  Changed several instances of PNG_NO_CONSOLE_ID to PNG_NO_STDIO in pngrutil.c
770    and mentioned the purposes of the two macros in libpng.txt/libpng.3.
771
772Version 1.0.6f [April 14, 2000]
773  Revised png_set_iCCP() and png_set_rows() to avoid prematurely freeing data.
774  Add checks in png_set_text() for NULL members of the input text structure.
775  Revised libpng.txt/libpng.3.
776  Removed superfluous prototype for png_set_iTXt from png.h
777  Removed "else" from pngread.c, after png_error(), and changed "0" to "length".
778  Changed several png_errors about malformed ancillary chunks to png_warnings.
779
780Version 1.0.6g [April 24, 2000]
781  Added png_pass-* arrays to pnggccrd.c when PNG_USE_LOCAL_ARRAYS is defined.
782  Relocated paragraph about png_set_background() in libpng.3/libpng.txt
783    and other revisions (Matthias Benckmann)
784  Relocated info_ptr->free_me, png_ptr->free_me, and other info_ptr and
785    png_ptr members to restore binary compatibility with libpng-1.0.5
786    (breaks compatibility with libpng-1.0.6).
787
788Version 1.0.6h [April 24, 2000]
789  Changed shared library so-number pattern from 2.x.y.z to xy.z (this builds
790    libpng.so.10 & libpng.so.10.6h instead of libpng.so.2 & libpng.so.2.1.0.6h)
791    This is a temporary change for test purposes.
792
793Version 1.0.6i [May 2, 2000]
794  Rearranged some members at the end of png_info and png_struct, to put
795    unknown_chunks_num and free_me within the original size of the png_structs
796    and free_me, png_read_user_fn, and png_free_fn within the original png_info,
797    because some old applications allocate the structs directly instead of
798    using png_create_*().
799  Added documentation of user memory functions in libpng.txt/libpng.3
800  Modified png_read_png so that it will use user_allocated row_pointers
801    if present, unless free_me directs that it be freed, and added description
802    of the use of png_set_rows() and png_get_rows() in libpng.txt/libpng.3.
803  Added PNG_LEGACY_SUPPORTED macro, and #ifdef out all new (since version
804    1.00) members of png_struct and png_info, to regain binary compatibility
805    when you define this macro.  Capabilities lost in this event
806    are user transforms (new in version 1.0.0),the user transform pointer
807    (new in version 1.0.2), rgb_to_gray (new in 1.0.5), iCCP, sCAL, sPLT,
808    the high-level interface, and unknown chunks support (all new in 1.0.6).
809    This was necessary because of old applications that allocate the structs
810    directly as authors were instructed to do in libpng-0.88 and earlier,
811    instead of using png_create_*().
812  Added modes PNG_CREATED_READ_STRUCT and PNG_CREATED_WRITE_STRUCT which
813    can be used to detect codes that directly allocate the structs, and
814    code to check these modes in png_read_init() and png_write_init() and
815    generate a libpng error if the modes aren't set and PNG_LEGACY_SUPPORTED
816    was not defined.
817  Added makefile.intel and updated makefile.watcom (Pawel Mrochen)
818
819Version 1.0.6j [May 3, 2000]
820  Overloaded png_read_init() and png_write_init() with macros that convert
821    calls to png_read_init_2() or png_write_init_2() that check the version
822    and structure sizes.
823
824Version 1.0.7beta11 [May 7, 2000]
825  Removed the new PNG_CREATED_READ_STRUCT and PNG_CREATED_WRITE_STRUCT modes
826    which are no longer used.
827  Eliminated the three new members of png_text when PNG_LEGACY_SUPPORTED is
828    defined or when neither PNG_READ_iTXt_SUPPORTED nor PNG_WRITE_iTXT_SUPPORTED
829    is defined.
830  Made PNG_NO_READ|WRITE_iTXt the default setting, to avoid memory
831    overrun when old applications fill the info_ptr->text structure directly.
832  Added PNGAPI macro, and added it to the definitions of all exported functions.
833  Relocated version macro definitions ahead of the includes of zlib.h and
834    pngconf.h in png.h.
835
836Version 1.0.7beta12 [May 12, 2000]
837  Revised pngset.c to avoid a problem with expanding the png_debug macro.
838  Deleted some extraneous defines from pngconf.h
839  Made PNG_NO_CONSOLE_IO the default condition when PNG_BUILD_DLL is defined.
840  Use MSC _RPTn debugging instead of fprintf if _MSC_VER is defined.
841  Added png_access_version_number() function.
842  Check for mask&PNG_FREE_CHNK (for TEXT, SCAL, PCAL) in png_free_data().
843  Expanded libpng.3/libpng.txt information about png_data_freer().
844
845Version 1.0.7beta14 [May 17, 2000] (beta13 was not published)
846  Changed pnggccrd.c and pngvcrd.c to handle bad adaptive filter types as
847    warnings instead of errors, as pngrutil.c does.
848  Set the PNG_INFO_IDAT valid flag in png_set_rows() so png_write_png()
849    will actually write IDATs.
850  Made the default PNG_USE_LOCAL_ARRAYS depend on PNG_DLL instead of WIN32.
851  Make png_free_data() ignore its final parameter except when freeing data
852    that can have multiple instances (text, sPLT, unknowns).
853  Fixed a new bug in png_set_rows().
854  Removed info_ptr->valid tests from png_free_data(), as in version 1.0.5.
855  Added png_set_invalid() function.
856  Fixed incorrect illustrations of png_destroy_write_struct() in example.c.
857
858Version 1.0.7beta15 [May 30, 2000]
859  Revised the deliberately erroneous Linux setjmp code in pngconf.h to produce
860    fewer error messages.
861  Rearranged checks for Z_OK to check the most likely path first in pngpread.c
862    and pngwutil.c.
863  Added checks in pngtest.c for png_create_*() returning NULL, and mentioned
864    in libpng.txt/libpng.3 the need for applications to check this.
865  Changed names of png_default_*() functions in pngtest to pngtest_*().
866  Changed return type of png_get_x|y_offset_*() from png_uint_32 to png_int_32.
867  Fixed some bugs in the unused PNG_INCH_CONVERSIONS functions in pngget.c
868  Set each pointer to NULL after freeing it in png_free_data().
869  Worked around a problem in pngconf.h; AIX's strings.h defines an "index"
870    macro that conflicts with libpng's png_color_16.index. (Dimitri
871    Papadapoulos)
872  Added "msvc" directory with MSVC++ project files (Simon-Pierre Cadieux).
873
874Version 1.0.7beta16 [June 4, 2000]
875  Revised the workaround of AIX string.h "index" bug.
876  Added a check for overlength PLTE chunk in pngrutil.c.
877  Added PNG_NO_POINTER_INDEXING macro to use array-indexing instead of pointer
878    indexing in pngrutil.c and pngwutil.c to accommodate a buggy compiler.
879  Added a warning in png_decompress_chunk() when it runs out of data, e.g.
880    when it tries to read an erroneous PhotoShop iCCP chunk.
881  Added PNG_USE_DLL macro.
882  Revised the copyright/disclaimer/license notice.
883  Added contrib/msvctest directory
884
885Version 1.0.7rc1 [June 9, 2000]
886  Corrected the definition of PNG_TRANSFORM_INVERT_ALPHA  (0x0400 not 0x0200)
887  Added contrib/visupng directory (Willem van Schaik)
888
889Version 1.0.7beta18 [June 23, 2000]
890  Revised PNGAPI definition, and pngvcrd.c to work with __GCC__
891    and do not redefine PNGAPI if it is passed in via a compiler directive.
892  Revised visupng/PngFile.c to remove returns from within the Try block.
893  Removed leading underscores from "_PNG_H" and "_PNG_SAVE_BSD_SOURCE" macros.
894  Updated contrib/visupng/cexcept.h to version 1.0.0.
895  Fixed bugs in pngwrite.c and pngwutil.c that prevented writing iCCP chunks.
896
897Version 1.0.7rc2 [June 28, 2000]
898  Updated license to include disclaimers required by UCITA.
899  Fixed "DJBPP" typo in pnggccrd.c introduced in beta18.
900
901Version 1.0.7 [July 1, 2000]
902  Revised the definition of "trans_values" in libpng.3/libpng.txt
903
904Version 1.0.8beta1 [July 8, 2000]
905  Added png_free(png_ptr, key) two places in pngpread.c to stop memory leaks.
906  Changed PNG_NO_STDIO to PNG_NO_CONSOLE_IO, several places in pngrutil.c and
907     pngwutil.c.
908  Changed PNG_EXPORT_VAR to use PNG_IMPEXP, in pngconf.h.
909  Removed unused "#include <assert.h>" from png.c
910  Added WindowsCE support.
911  Revised pnggccrd.c to work with gcc-2.95.2 and in the Cygwin environment.
912
913Version 1.0.8beta2 [July 10, 2000]
914  Added project files to the wince directory and made further revisions
915     of pngtest.c, pngrio.c, and pngwio.c in support of WindowsCE.
916
917Version 1.0.8beta3 [July 11, 2000]
918  Only set the PNG_FLAG_FREE_TRNS or PNG_FREE_TRNS flag in png_handle_tRNS()
919     for indexed-color input files to avoid potential double-freeing trans array
920     under some unusual conditions; problem was introduced in version 1.0.6f.
921  Further revisions to pngtest.c and files in the wince subdirectory.
922
923Version 1.0.8beta4 [July 14, 2000]
924  Added the files pngbar.png and pngbar.jpg to the distribution.
925  Added makefile.cygwin, and cygwin support in pngconf.h
926  Added PNG_NO_ZALLOC_ZERO macro (makes png_zalloc skip zeroing memory)
927
928Version 1.0.8rc1 [July 16, 2000]
929  Revised png_debug() macros and statements to eliminate compiler warnings.
930
931Version 1.0.8 [July 24, 2000]
932  Added png_flush() in pngwrite.c, after png_write_IEND().
933  Updated makefile.hpux to build a shared library.
934
935Version 1.0.9beta1 [November 10, 2000]
936  Fixed typo in scripts/makefile.hpux
937  Updated makevms.com in scripts and contrib/* and contrib/* (Martin Zinser)
938  Fixed seqence-point bug in contrib/pngminus/png2pnm (Martin Zinser)
939  Changed "cdrom.com" in documentation to "libpng.org"
940  Revised pnggccrd.c to get it all working, and updated makefile.gcmmx (Greg).
941  Changed type of "params" from voidp to png_voidp in png_read|write_png().
942  Make sure PNGAPI and PNG_IMPEXP are defined in pngconf.h.
943  Revised the 3 instances of WRITEFILE in pngtest.c.
944  Relocated "msvc" and "wince" project subdirectories into "dll" subdirectory.
945  Updated png.rc in dll/msvc project
946  Revised makefile.dec to define and use LIBPATH and INCPATH
947  Increased size of global png_libpng_ver[] array from 12 to 18 chars.
948  Made global png_libpng_ver[], png_sig[] and png_pass_*[] arrays const.
949  Removed duplicate png_crc_finish() from png_handle_bKGD() function.
950  Added a warning when application calls png_read_update_info() multiple times.
951  Revised makefile.cygwin
952  Fixed bugs in iCCP support in pngrutil.c and pngwutil.c.
953  Replaced png_set_empty_plte_permitted() with png_permit_mng_features().
954
955Version 1.0.9beta2 [November 19, 2000]
956  Renamed the "dll" subdirectory "projects".
957  Added borland project files to "projects" subdirectory.
958  Set VS_FF_PRERELEASE and VS_FF_PATCHED flags in msvc/png.rc when appropriate.
959  Add error message in png_set_compression_buffer_size() when malloc fails.
960
961Version 1.0.9beta3 [November 23, 2000]
962  Revised PNG_LIBPNG_BUILD_TYPE macro in png.h, used in the msvc project.
963  Removed the png_flush() in pngwrite.c that crashes some applications
964    that don't set png_output_flush_fn.
965  Added makefile.macosx and makefile.aix to scripts directory.
966
967Version 1.0.9beta4 [December 1, 2000]
968  Change png_chunk_warning to png_warning in png_check_keyword().
969  Increased the first part of msg buffer from 16 to 18 in png_chunk_error().
970
971Version 1.0.9beta5 [December 15, 2000]
972  Added support for filter method 64 (for PNG datastreams embedded in MNG).
973
974Version 1.0.9beta6 [December 18, 2000]
975  Revised png_set_filter() to accept filter method 64 when appropriate.
976  Added new PNG_HAVE_PNG_SIGNATURE bit to png_ptr->mode and use it to
977    help prevent applications from using MNG features in PNG datastreams.
978  Added png_permit_mng_features() function.
979  Revised libpng.3/libpng.txt.  Changed "filter type" to "filter method".
980
981Version 1.0.9rc1 [December 23, 2000]
982  Revised test for PNG_HAVE_PNG_SIGNATURE in pngrutil.c
983  Fixed error handling of unknown compression type in png_decompress_chunk().
984  In pngconf.h, define __cdecl when _MSC_VER is defined.
985
986Version 1.0.9beta7 [December 28, 2000]
987  Changed PNG_TEXT_COMPRESSION_zTXt to PNG_COMPRESSION_TYPE_BASE several places.
988  Revised memory management in png_set_hIST and png_handle_hIST in a backward
989    compatible manner.  PLTE and tRNS were revised similarly.
990  Revised the iCCP chunk reader to ignore trailing garbage.
991
992Version 1.0.9beta8 [January 12, 2001]
993  Moved pngasmrd.h into pngconf.h.
994  Improved handling of out-of-spec garbage iCCP chunks generated by PhotoShop.
995
996Version 1.0.9beta9 [January 15, 2001]
997  Added png_set_invalid, png_permit_mng_features, and png_mmx_supported to
998    wince and msvc project module definition files.
999  Minor revision of makefile.cygwin.
1000  Fixed bug with progressive reading of narrow interlaced images in pngpread.c
1001
1002Version 1.0.9beta10 [January 16, 2001]
1003  Do not typedef png_FILE_p in pngconf.h when PNG_NO_STDIO is defined.
1004  Fixed "png_mmx_supported" typo in project definition files.
1005
1006Version 1.0.9beta11 [January 19, 2001]
1007  Updated makefile.sgi to make shared library.
1008  Removed png_mmx_support() function and disabled PNG_MNG_FEATURES_SUPPORTED
1009    by default, for the benefit of DLL forward compatibility.  These will
1010    be re-enabled in version 1.2.0.
1011
1012Version 1.0.9rc2 [January 22, 2001]
1013  Revised cygwin support.
1014
1015Version 1.0.9 [January 31, 2001]
1016  Added check of cygwin's ALL_STATIC in pngconf.h
1017  Added "-nommx" parameter to contrib/gregbook/rpng2-win and rpng2-x demos.
1018
1019Version 1.0.10beta1 [March 14, 2001]
1020  Revised makefile.dec, makefile.sgi, and makefile.sggcc; added makefile.hpgcc.
1021  Reformatted libpng.3 to eliminate bad line breaks.
1022  Added checks for _mmx_supported in the read_filter_row function of pnggccrd.c
1023  Added prototype for png_mmx_support() near the top of pnggccrd.c
1024  Moved some error checking from png_handle_IHDR to png_set_IHDR.
1025  Added PNG_NO_READ_SUPPORTED and PNG_NO_WRITE_SUPPORTED macros.
1026  Revised png_mmx_support() function in pnggccrd.c
1027  Restored version 1.0.8 PNG_WRITE_EMPTY_PLTE_SUPPORTED behavior in pngwutil.c
1028  Fixed memory leak in contrib/visupng/PngFile.c
1029  Fixed bugs in png_combine_row() in pnggccrd.c and pngvcrd.c (C version)
1030  Added warnings when retrieving or setting gamma=0.
1031  Increased the first part of msg buffer from 16 to 18 in png_chunk_warning().
1032
1033Version 1.0.10rc1 [March 23, 2001]
1034  Changed all instances of memcpy, strcpy, and strlen to png_memcpy, png_strcpy,
1035    and png_strlen.
1036  Revised png_mmx_supported() function in pnggccrd.c to return proper value.
1037  Fixed bug in progressive reading (pngpread.c) with small images (height < 8).
1038
1039Version 1.0.10 [March 30, 2001]
1040  Deleted extraneous space (introduced in 1.0.9) from line 42 of makefile.cygwin
1041  Added beos project files (Chris Herborth)
1042
1043Version 1.0.11beta1 [April 3, 2001]
1044  Added type casts on several png_malloc() calls (Dimitri Papadapoulos).
1045  Removed a no-longer needed AIX work-around from pngconf.h
1046  Changed several "//" single-line comments to C-style in pnggccrd.c
1047
1048Version 1.0.11beta2 [April 11, 2001]
1049  Removed PNGAPI from several functions whose prototypes did not have PNGAPI.
1050  Updated scripts/pngos2.def
1051
1052Version 1.0.11beta3 [April 14, 2001]
1053  Added checking the results of many instances of png_malloc() for NULL
1054
1055Version 1.0.11beta4 [April 20, 2001]
1056  Undid the changes from version 1.0.11beta3.  Added a check for NULL return
1057    from user's malloc_fn().
1058  Removed some useless type casts of the NULL pointer.
1059  Added makefile.netbsd
1060
1061Version 1.0.11 [April 27, 2001]
1062  Revised makefile.netbsd
1063
1064Version 1.0.12beta1 [May 14, 2001]
1065  Test for Windows platform in pngconf.h when including malloc.h (Emmanuel Blot)
1066  Updated makefile.cygwin and handling of Cygwin's ALL_STATIC in pngconf.h
1067  Added some never-to-be-executed code in pnggccrd.c to quiet compiler warnings.
1068  Eliminated the png_error about apps using png_read|write_init().  Instead,
1069    libpng will reallocate the png_struct and info_struct if they are too small.
1070    This retains future binary compatibility for old applications written for
1071    libpng-0.88 and earlier.
1072
1073Version 1.2.0beta1 [May 6, 2001]
1074  Bumped DLLNUM to 2.
1075  Re-enabled PNG_MNG_FEATURES_SUPPORTED and enabled PNG_ASSEMBLER_CODE_SUPPORTED
1076    by default.
1077  Added runtime selection of MMX features.
1078  Added png_set_strip_error_numbers function and related macros.
1079
1080Version 1.2.0beta2 [May 7, 2001]
1081  Finished merging 1.2.0beta1 with version 1.0.11
1082  Added a check for attempts to read or write PLTE in grayscale PNG datastreams.
1083
1084Version 1.2.0beta3 [May 17, 2001]
1085  Enabled user memory function by default.
1086  Modified png_create_struct so it passes user mem_ptr to user memory allocator.
1087  Increased png_mng_features flag from png_byte to png_uint_32.
1088  Bumped shared-library (so-number) and dll-number to 3.
1089
1090Version 1.2.0beta4 [June 23, 2001]
1091  Check for missing profile length field in iCCP chunk and free chunk_data
1092     in case of truncated iCCP chunk.
1093  Bumped shared-library number to 3 in makefile.sgi and makefile.sggcc
1094  Bumped dll-number from 2 to 3 in makefile.cygwin
1095  Revised contrib/gregbook/rpng*-x.c to avoid a memory leak and to exit cleanly
1096     if user attempts to run it on an 8-bit display.
1097  Updated contrib/gregbook
1098  Use png_malloc instead of png_zalloc to allocate palette in pngset.c
1099  Updated makefile.ibmc
1100  Added some typecasts to eliminate gcc 3.0 warnings.  Changed prototypes
1101     of png_write_oFFS width and height from png_uint_32 to png_int_32.
1102  Updated example.c
1103  Revised prototypes for png_debug_malloc and png_debug_free in pngtest.c
1104
1105Version 1.2.0beta5 [August 8, 2001]
1106  Revised contrib/gregbook
1107  Revised makefile.gcmmx
1108  Revised pnggccrd.c to conditionally compile some thread-unsafe code only
1109     when PNG_THREAD_UNSAFE_OK is defined.
1110  Added tests to prevent pngwutil.c from writing a bKGD or tRNS chunk with
1111     value exceeding 2^bit_depth-1
1112  Revised makefile.sgi and makefile.sggcc
1113  Replaced calls to fprintf(stderr,...) with png_warning() in pnggccrd.c
1114  Removed restriction that do_invert_mono only operate on 1-bit opaque files
1115
1116Version 1.2.0 [September 1, 2001]
1117  Changed a png_warning() to png_debug() in pnggccrd.c
1118  Fixed contrib/gregbook/rpng-x.c, rpng2-x.c to avoid crash with XFreeGC().
1119
1120Version 1.2.1beta1 [October 19, 2001]
1121  Revised makefile.std in contrib/pngminus
1122  Include background_1 in png_struct regardless of gamma support.
1123  Revised makefile.netbsd and makefile.macosx, added makefile.darwin.
1124  Revised example.c to provide more details about using row_callback().
1125
1126Version 1.2.1beta2 [October 25, 2001]
1127  Added type cast to each NULL appearing in a function call, except for
1128    WINCE functions.
1129  Added makefile.so9.
1130
1131Version 1.2.1beta3 [October 27, 2001]
1132  Removed type casts from all NULLs.
1133  Simplified png_create_struct_2().
1134
1135Version 1.2.1beta4 [November 7, 2001]
1136  Revised png_create_info_struct() and png_creat_struct_2().
1137  Added error message if png_write_info() was omitted.
1138  Type cast NULLs appearing in function calls when _NO_PROTO or
1139    PNG_TYPECAST_NULL is defined.
1140
1141Version 1.2.1rc1 [November 24, 2001]
1142  Type cast NULLs appearing in function calls except when PNG_NO_TYPECAST_NULL
1143    is defined.
1144  Changed typecast of "size" argument to png_size_t in pngmem.c calls to
1145    the user malloc_fn, to agree with the prototype in png.h
1146  Added a pop/push operation to pnggccrd.c, to preserve Eflag (Maxim Sobolev)
1147  Updated makefile.sgi to recognize LIBPATH and INCPATH.
1148  Updated various makefiles so "make clean" does not remove previous major
1149    version of the shared library.
1150
1151Version 1.2.1rc2 [December 4, 2001]
1152  Always allocate 256-entry internal palette, hist, and trans arrays, to
1153    avoid out-of-bounds memory reference caused by invalid PNG datastreams.
1154  Added a check for prefix_length > data_length in iCCP chunk handler.
1155
1156Version 1.2.1 [December 7, 2001]
1157  None.
1158
1159Version 1.2.2beta1 [February 22, 2002]
1160  Fixed a bug with reading the length of iCCP profiles (Larry Reeves).
1161  Revised makefile.linux, makefile.gcmmx, and makefile.sgi to generate
1162    libpng.a, libpng12.so (not libpng.so.3), and libpng12/png.h
1163  Revised makefile.darwin to remove "-undefined suppress" option.
1164  Added checks for gamma and chromaticity values over 21474.83, which exceed
1165    the limit for PNG unsigned 32-bit integers when encoded.
1166  Revised calls to png_create_read_struct() and png_create_write_struct()
1167    for simpler debugging.
1168  Revised png_zalloc() so zlib handles errors (uses PNG_FLAG_MALLOC_NULL_MEM_OK)
1169
1170Version 1.2.2beta2 [February 23, 2002]
1171  Check chunk_length and idat_size for invalid (over PNG_MAX_UINT) lengths.
1172  Check for invalid image dimensions in png_get_IHDR.
1173  Added missing "fi;" in the install target of the SGI makefiles.
1174  Added install-static to all makefiles that make shared libraries.
1175  Always do gamma compensation when image is partially transparent.
1176
1177Version 1.2.2beta3 [March 7, 2002]
1178  Compute background.gray and background_1.gray even when color_type is RGB
1179    in case image gets reduced to gray later.
1180  Modified shared-library makefiles to install pkgconfig/libpngNN.pc.
1181  Export (with PNGAPI) png_zalloc, png_zfree, and png_handle_as_unknown
1182  Removed unused png_write_destroy_info prototype from png.h
1183  Eliminated incorrect use of width_mmx from pnggccrd.c in pixel_bytes == 8 case
1184  Added install-shared target to all makefiles that make shared libraries.
1185  Stopped a double free of palette, hist, and trans when not using free_me.
1186  Added makefile.32sunu for Sun Ultra 32 and makefile.64sunu for Sun Ultra 64.
1187
1188Version 1.2.2beta4 [March 8, 2002]
1189  Compute background.gray and background_1.gray even when color_type is RGB
1190    in case image gets reduced to gray later (Jason Summers).
1191  Relocated a misplaced /bin/rm in the "install-shared" makefile targets
1192  Added PNG_1_0_X macro which can be used to build a 1.0.x-compatible library.
1193
1194Version 1.2.2beta5 [March 26, 2002]
1195  Added missing PNGAPI to several function definitions.
1196  Check for invalid bit_depth or color_type in png_get_IHDR(), and
1197    check for missing PLTE or IHDR in png_push_read_chunk() (Matthias Clasen).
1198  Revised iTXt support to accept NULL for lang and lang_key.
1199  Compute gamma for color components of background even when color_type is gray.
1200  Changed "()" to "{}" in scripts/libpng.pc.in.
1201  Revised makefiles to put png.h and pngconf.h only in $prefix/include/libpngNN
1202  Revised makefiles to make symlink to libpng.so.NN in addition to libpngNN.so
1203
1204Version 1.2.2beta6 [March 31, 2002]
1205
1206Version 1.0.13beta1 [March 31, 2002]
1207  Prevent png_zalloc() from trying to memset memory that it failed to acquire.
1208  Add typecasts of PNG_MAX_UINT in pngset_cHRM_fixed() (Matt Holgate).
1209  Ensure that the right function (user or default) is used to free the
1210    png_struct after an error in png_create_read_struct_2().
1211
1212Version 1.2.2rc1 [April 7, 2002]
1213
1214Version 1.0.13rc1 [April 7, 2002]
1215  Save the ebx register in pnggccrd.c (Sami Farin)
1216  Add "mem_ptr = png_ptr->mem_ptr" in png_destroy_write_struct() (Paul Gardner).
1217  Updated makefiles to put headers in include/libpng and remove old include/*.h.
1218
1219Version 1.2.2 [April 15, 2002]
1220
1221Version 1.0.13 [April 15, 2002]
1222  Revised description of png_set_filter() in libpng.3/libpng.txt.
1223  Revised makefile.netbsd and added makefile.neNNbsd and makefile.freebsd
1224
1225Version 1.0.13patch01 [April 17, 2002]
1226
1227Version 1.2.2patch01 [April 17, 2002]
1228  Changed ${PNGMAJ}.${PNGVER} bug to ${PNGVER} in makefile.sgi and
1229    makefile.sggcc
1230  Fixed VER -> PNGVER typo in makefile.macosx and added install-static to
1231    install
1232  Added install: target to makefile.32sunu and makefile.64sunu
1233
1234Version 1.0.13patch03 [April 18, 2002]
1235
1236Version 1.2.2patch03 [April 18, 2002]
1237  Revised 15 makefiles to link libpng.a to libpngNN.a and the include libpng
1238  subdirectory to libpngNN subdirectory without the full pathname.
1239  Moved generation of libpng.pc from "install" to "all" in 15 makefiles.
1240
1241Version 1.2.3rc1 [April 28, 2002]
1242  Added install-man target to 15 makefiles (Dimitri Papadopolous-Orfanos).
1243  Added $(DESTDIR) feature to 24 makefiles (Tim Mooney)
1244  Fixed bug with $prefix, should be $(prefix) in makefile.hpux.
1245  Updated cygwin-specific portion of pngconf.h and revised makefile.cygwin
1246  Added a link from libpngNN.pc to libpng.pc in 15 makefiles.
1247  Added links from include/libpngNN/*.h to include/*.h in 24 makefiles.
1248  Revised makefile.darwin to make relative links without full pathname.
1249  Added setjmp() at the end of png_create_*_struct_2() in case user forgets
1250    to put one in their application.
1251  Restored png_zalloc() and png_zfree() prototypes to version 1.2.1 and
1252    removed them from module definition files.
1253
1254Version 1.2.3rc2 [May 1, 2002]
1255  Fixed bug in reporting number of channels in pngget.c and pngset.c,
1256    that was introduced in version 1.2.2beta5.
1257  Exported png_zalloc(), png_zfree(), png_default_read(), png_default_write(),
1258    png_default_flush(), and png_push_fill_buffer() and included them in
1259    module definition files.
1260  Added "libpng.pc" dependency to the "install-shared" target in 15 makefiles.
1261
1262Version 1.2.3rc3 [May 1, 2002]
1263  Revised prototype for png_default_flush()
1264  Remove old libpng.pc and libpngNN.pc before installing new ones.
1265
1266Version 1.2.3rc4 [May 2, 2002]
1267  Typos in *.def files (png_default_read|write -> png_default_read|write_data)
1268  In makefiles, changed rm libpng.NN.pc to rm libpngNN.pc
1269  Added libpng-config and libpngNN-config and modified makefiles to install
1270    them.
1271  Changed $(MANPATH) to $(DESTDIR)$(MANPATH) in makefiles
1272  Added "Win32 DLL VB" configuration to projects/msvc/libpng.dsp
1273
1274Version 1.2.3rc5 [May 11, 2002]
1275  Changed "error" and "message" in prototypes to "error_message" and
1276    "warning_message" to avoid namespace conflict.
1277  Revised 15 makefiles to build libpng-config from libpng-config-*.in
1278  Once more restored png_zalloc and png_zfree to regular nonexported form.
1279  Restored png_default_read|write_data, png_default_flush, png_read_fill_buffer
1280    to nonexported form, but with PNGAPI, and removed them from module def
1281    files.
1282
1283Version 1.2.3rc6 [May 14, 2002]
1284  Removed "PNGAPI" from png_zalloc() and png_zfree() in png.c
1285  Changed "Gz" to "Gd" in projects/msvc/libpng.dsp and zlib.dsp.
1286  Removed leftover libpng-config "sed" script from four makefiles.
1287  Revised libpng-config creating script in 16 makefiles.
1288
1289Version 1.2.3 [May 22, 2002]
1290  Revised libpng-config target in makefile.cygwin.
1291  Removed description of png_set_mem_fn() from documentation.
1292  Revised makefile.freebsd.
1293  Minor cosmetic changes to 15 makefiles, e.g., $(DI) = $(DESTDIR)/$(INCDIR).
1294  Revised projects/msvc/README.txt
1295  Changed -lpng to -lpngNN in LDFLAGS in several makefiles.
1296
1297Version 1.2.4beta1 [May 24, 2002]
1298  Added libpng.pc and libpng-config to "all:" target in 16 makefiles.
1299  Fixed bug in 16 makefiles: $(DESTDIR)/$(LIBPATH) to $(DESTDIR)$(LIBPATH)
1300  Added missing "\" before closing double quote in makefile.gcmmx.
1301  Plugged various memory leaks; added png_malloc_warn() and png_set_text_2()
1302    functions.
1303
1304Version 1.2.4beta2 [June 25, 2002]
1305  Plugged memory leak of png_ptr->current_text (Matt Holgate).
1306  Check for buffer overflow before reading CRC in pngpread.c (Warwick Allison)
1307  Added -soname to the loader flags in makefile.dec, makefile.sgi, and
1308    makefile.sggcc.
1309  Added "test-installed" target to makefile.linux, makefile.gcmmx,
1310    makefile.sgi, and makefile.sggcc.
1311
1312Version 1.2.4beta3 [June 28, 2002]
1313  Plugged memory leak of row_buf in pngtest.c when there is a png_error().
1314  Detect buffer overflow in pngpread.c when IDAT is corrupted with extra data.
1315  Added "test-installed" target to makefile.32sunu, makefile.64sunu,
1316    makefile.beos, makefile.darwin, makefile.dec, makefile.macosx,
1317    makefile.solaris, makefile.hpux, makefile.hpgcc, and makefile.so9.
1318
1319Version 1.2.4rc1 and 1.0.14rc1 [July 2, 2002]
1320  Added "test-installed" target to makefile.cygwin and makefile.sco.
1321  Revised pnggccrd.c to be able to back out version 1.0.x via PNG_1_0_X macro.
1322
1323Version 1.2.4 and 1.0.14 [July 8, 2002]
1324  Changed png_warning() to png_error() when width is too large to process.
1325
1326Version 1.2.4patch01 [July 20, 2002]
1327  Revised makefile.cygwin to use DLL number 12 instead of 13.
1328
1329Version 1.2.5beta1 [August 6, 2002]
1330  Added code to contrib/gregbook/readpng2.c to ignore unused chunks.
1331  Replaced toucan.png in contrib/gregbook (it has been corrupt since 1.0.11)
1332  Removed some stray *.o files from contrib/gregbook.
1333  Changed png_error() to png_warning() about "Too much data" in pngpread.c
1334    and about "Extra compressed data" in pngrutil.c.
1335  Prevent png_ptr->pass from exceeding 7 in png_push_finish_row().
1336  Updated makefile.hpgcc
1337  Updated png.c and pnggccrd.c handling of return from png_mmx_support()
1338
1339Version 1.2.5beta2 [August 15, 2002]
1340  Only issue png_warning() about "Too much data" in pngpread.c when avail_in
1341    is nonzero.
1342  Updated makefiles to install a separate libpng.so.3 with its own rpath.
1343
1344Version 1.2.5rc1 and 1.0.15rc1 [August 24, 2002]
1345  Revised makefiles to not remove previous minor versions of shared libraries.
1346
1347Version 1.2.5rc2 and 1.0.15rc2 [September 16, 2002]
1348  Revised 13 makefiles to remove "-lz" and "-L$(ZLIBLIB)", etc., from shared
1349    library loader directive.
1350  Added missing "$OBJSDLL" line to makefile.gcmmx.
1351  Added missing "; fi" to makefile.32sunu.
1352
1353Version 1.2.5rc3 and 1.0.15rc3 [September 18, 2002]
1354  Revised libpng-config script.
1355
1356Version 1.2.5 and 1.0.15 [October 3, 2002]
1357  Revised makefile.macosx, makefile.darwin, makefile.hpgcc, and makefile.hpux,
1358    and makefile.aix.
1359  Relocated two misplaced PNGAPI lines in pngtest.c
1360
1361Version 1.2.6beta1 [October 22, 2002]
1362  Commented out warning about uninitialized mmx_support in pnggccrd.c.
1363  Changed "IBMCPP__" flag to "__IBMCPP__" in pngconf.h.
1364  Relocated two more misplaced PNGAPI lines in pngtest.c
1365  Fixed memory overrun bug in png_do_read_filler() with 16-bit datastreams,
1366    introduced in version 1.0.2.
1367  Revised makefile.macosx, makefile.dec, makefile.aix, and makefile.32sunu.
1368
1369Version 1.2.6beta2 [November 1, 2002]
1370  Added libpng-config "--ldopts" output.
1371  Added "AR=ar" and "ARFLAGS=rc" and changed "ar rc" to "$(AR) $(ARFLAGS)"
1372    in makefiles.
1373
1374Version 1.2.6beta3 [July 18, 2004]
1375  Reverted makefile changes from version 1.2.6beta2 and some of the changes
1376    from version 1.2.6beta1; these will be postponed until version 1.2.7.
1377    Version 1.2.6 is going to be a simple bugfix release.
1378  Changed the one instance of "ln -sf" to "ln -f -s" in each Sun makefile.
1379  Fixed potential overrun in pngerror.c by using strncpy instead of memcpy.
1380  Added "#!/bin/sh" at the top of configure, for recognition of the
1381    'x' flag under Cygwin (Cosmin).
1382  Optimized vacuous tests that silence compiler warnings, in png.c (Cosmin).
1383  Added support for PNG_USER_CONFIG, in pngconf.h (Cosmin).
1384  Fixed the special memory handler for Borland C under DOS, in pngmem.c
1385    (Cosmin).
1386  Removed some spurious assignments in pngrutil.c (Cosmin).
1387  Replaced 65536 with 65536L, and 0xffff with 0xffffL, to silence warnings
1388    on 16-bit platforms (Cosmin).
1389  Enclosed shift op expressions in parentheses, to silence warnings (Cosmin).
1390  Used proper type png_fixed_point, to avoid problems on 16-bit platforms,
1391    in png_handle_sRGB() (Cosmin).
1392  Added compression_type to png_struct, and optimized the window size
1393    inside the deflate stream (Cosmin).
1394  Fixed definition of isnonalpha(), in pngerror.c and pngrutil.c (Cosmin).
1395  Fixed handling of unknown chunks that come after IDAT (Cosmin).
1396  Allowed png_error() and png_warning() to work even if png_ptr == NULL
1397    (Cosmin).
1398  Replaced row_info->rowbytes with row_bytes in png_write_find_filter()
1399    (Cosmin).
1400  Fixed definition of PNG_LIBPNG_VER_DLLNUM (Simon-Pierre).
1401  Used PNG_LIBPNG_VER and PNG_LIBPNG_VER_STRING instead of the hardcoded
1402    values in png.c (Simon-Pierre, Cosmin).
1403  Initialized png_libpng_ver[] with PNG_LIBPNG_VER_STRING (Simon-Pierre).
1404  Replaced PNG_LIBPNG_VER_MAJOR with PNG_LIBPNG_VER_DLLNUM in png.rc
1405    (Simon-Pierre).
1406  Moved the definition of PNG_HEADER_VERSION_STRING near the definitions
1407    of the other PNG_LIBPNG_VER_... symbols in png.h (Cosmin).
1408  Relocated #ifndef PNGAPI guards in pngconf.h (Simon-Pierre, Cosmin).
1409  Updated scripts/makefile.vc(a)win32 (Cosmin).
1410  Updated the MSVC project (Simon-Pierre, Cosmin).
1411  Updated the Borland C++ Builder project (Cosmin).
1412  Avoided access to asm_flags in pngvcrd.c, if PNG_1_0_X is defined (Cosmin).
1413  Commented out warning about uninitialized mmx_support in pngvcrd.c (Cosmin).
1414  Removed scripts/makefile.bd32 and scripts/pngdef.pas (Cosmin).
1415  Added extra guard around inclusion of Turbo C memory headers, in pngconf.h
1416    (Cosmin).
1417  Renamed projects/msvc/ to projects/visualc6/, and projects/borland/ to
1418    projects/cbuilder5/ (Cosmin).
1419  Moved projects/visualc6/png32ms.def to scripts/pngw32.def,
1420    and projects/visualc6/png.rc to scripts/pngw32.rc (Cosmin).
1421  Added projects/visualc6/pngtest.dsp; removed contrib/msvctest/ (Cosmin).
1422  Changed line endings to DOS style in cbuilder5 and visualc6 files, even
1423    in the tar.* distributions (Cosmin).
1424  Updated contrib/visupng/VisualPng.dsp (Cosmin).
1425  Updated contrib/visupng/cexcept.h to version 2.0.0 (Cosmin).
1426  Added a separate distribution with "configure" and supporting files (Junichi).
1427
1428Version 1.2.6beta4 [July 28, 2004]
1429  Added user ability to change png_size_t via a PNG_SIZE_T macro.
1430  Added png_sizeof() and png_convert_size() functions.
1431  Added PNG_SIZE_MAX (maximum value of a png_size_t variable.
1432  Added check in png_malloc_default() for (size_t)size != (png_uint_32)size
1433    which would indicate an overflow.
1434  Changed sPLT failure action from png_error to png_warning and abandon chunk.
1435  Changed sCAL and iCCP failures from png_error to png_warning and abandon.
1436  Added png_get_uint_31(png_ptr, buf) function.
1437  Added PNG_UINT_32_MAX macro.
1438  Renamed PNG_MAX_UINT to PNG_UINT_31_MAX.
1439  Made png_zalloc() issue a png_warning and return NULL on potential
1440    overflow.
1441  Turn on PNG_NO_ZALLOC_ZERO by default in version 1.2.x
1442  Revised "clobber list" in pnggccrd.c so it will compile under gcc-3.4.
1443  Revised Borland portion of png_malloc() to return NULL or issue
1444    png_error() according to setting of PNG_FLAG_MALLOC_NULL_MEM_OK.
1445  Added PNG_NO_SEQUENTIAL_READ_SUPPORTED macro to conditionally remove
1446    sequential read support.
1447  Added some "#if PNG_WRITE_SUPPORTED" blocks.
1448  Added #ifdef to remove some redundancy in png_malloc_default().
1449  Use png_malloc instead of png_zalloc to allocate the pallete.
1450
1451Version 1.0.16rc1 and 1.2.6rc1 [August 4, 2004]
1452  Fixed buffer overflow vulnerability in png_handle_tRNS()
1453  Fixed integer arithmetic overflow vulnerability in png_read_png().
1454  Fixed some harmless bugs in png_handle_sBIT, etc, that would cause
1455    duplicate chunk types to go undetected.
1456  Fixed some timestamps in the -config version
1457  Rearranged order of processing of color types in png_handle_tRNS().
1458  Added ROWBYTES macro to calculate rowbytes without integer overflow.
1459  Updated makefile.darwin and removed makefile.macosx from scripts directory.
1460  Imposed default one million column, one-million row limits on the image
1461    dimensions, and added png_set_user_limits() function to override them.
1462  Revised use of PNG_SET_USER_LIMITS_SUPPORTED macro.
1463  Fixed wrong cast of returns from png_get_user_width|height_max().
1464  Changed some "keep the compiler happy" from empty statements to returns,
1465  Revised libpng.txt to remove 1.2.x stuff from the 1.0.x distribution
1466
1467Version 1.0.16rc2 and 1.2.6rc2 [August 7, 2004]
1468  Revised makefile.darwin and makefile.solaris.  Removed makefile.macosx.
1469  Revised pngtest's png_debug_malloc() to use png_malloc() instead of
1470    png_malloc_default() which is not supposed to be exported.
1471  Fixed off-by-one error in one of the conversions to PNG_ROWBYTES() in
1472    pngpread.c.  Bug was introduced in 1.2.6rc1.
1473  Fixed bug in RGB to RGBX transformation introduced in 1.2.6rc1.
1474  Fixed old bug in RGB to Gray transformation.
1475  Fixed problem with 64-bit compilers by casting arguments to abs()
1476    to png_int_32.
1477  Changed "ln -sf" to "ln -f -s" in three makefiles (solaris, sco, so9).
1478  Changed "HANDLE_CHUNK_*" to "PNG_HANDLE_CHUNK_*" (Cosmin)
1479  Added "-@/bin/rm -f $(DL)/$(LIBNAME).so.$(PNGMAJ)" to 15 *NIX makefiles.
1480  Added code to update the row_info->colortype in png_do_read_filler() (MSB).
1481
1482Version 1.0.16rc3 and 1.2.6rc3 [August 9, 2004]
1483  Eliminated use of "abs()" in testing cHRM and gAMA values, to avoid
1484    trouble with some 64-bit compilers.  Created PNG_OUT_OF_RANGE() macro.
1485  Revised documentation of png_set_keep_unknown_chunks().
1486  Check handle_as_unknown status in pngpread.c, as in pngread.c previously.
1487  Moved  "PNG_HANDLE_CHUNK_*" macros out of PNG_INTERNAL section of png.h
1488  Added "rim" definitions for CONST4 and CONST6 in pnggccrd.c
1489
1490Version 1.0.16rc4 and 1.2.6rc4 [August 10, 2004]
1491  Fixed mistake in pngtest.c introduced in 1.2.6rc2 (declaration of
1492    "pinfo" was out of place).
1493
1494Version 1.0.16rc5 and 1.2.6rc5 [August 10, 2004]
1495  Moved  "PNG_HANDLE_CHUNK_*" macros out of PNG_ASSEMBLER_CODE_SUPPORTED
1496     section of png.h where they were inadvertently placed in version rc3.
1497
1498Version 1.2.6 and 1.0.16 [August 15, 2004]
1499  Revised pngtest so memory allocation testing is only done when PNG_DEBUG==1.
1500
1501Version 1.2.7beta1 [August 26, 2004]
1502  Removed unused pngasmrd.h file.
1503  Removed references to uu.net for archived files.  Added references to
1504    PNG Spec (second edition) and the PNG ISO/IEC Standard.
1505  Added "test-dd" target in 15 makefiles, to run pngtest in DESTDIR.
1506  Fixed bug with "optimized window size" in the IDAT datastream, that
1507    causes libpng to write PNG files with incorrect zlib header bytes.
1508
1509Version 1.2.7beta2 [August 28, 2004]
1510  Fixed bug with sCAL chunk and big-endian machines (David Munro).
1511  Undid new code added in 1.2.6rc2 to update the color_type in
1512    png_set_filler().
1513  Added png_set_add_alpha() that updates color type.
1514
1515Version 1.0.17rc1 and 1.2.7rc1 [September 4, 2004]
1516  Revised png_set_strip_filler() to not remove alpha if color_type has alpha.
1517
1518Version 1.2.7 and 1.0.17 [September 12, 2004]
1519  Added makefile.hp64
1520  Changed projects/msvc/png32ms.def to scripts/png32ms.def in makefile.cygwin
1521
1522Version 1.2.8beta1 [November 1, 2004]
1523  Fixed bug in png_text_compress() that would fail to complete a large block.
1524  Fixed bug, introduced in libpng-1.2.7, that overruns a buffer during
1525    strip alpha operation in png_do_strip_filler().
1526  Added PNG_1_2_X definition in pngconf.h
1527  Use #ifdef to comment out png_info_init in png.c and png_read_init in
1528    pngread.c (as of 1.3.0)
1529
1530Version 1.2.8beta2 [November 2, 2004]
1531  Reduce color_type to a nonalpha type after strip alpha operation in
1532    png_do_strip_filler().
1533
1534Version 1.2.8beta3 [November 3, 2004]
1535  Revised definitions of PNG_MAX_UINT_32, PNG_MAX_SIZE, and PNG_MAXSUM
1536
1537Version 1.2.8beta4 [November 12, 2004]
1538  Fixed (again) definition of PNG_LIBPNG_VER_DLLNUM in png.h (Cosmin).
1539  Added PNG_LIBPNG_BUILD_PRIVATE in png.h (Cosmin).
1540  Set png_ptr->zstream.data_type to Z_BINARY, to avoid unnecessary detection
1541    of data type in deflate (Cosmin).
1542  Deprecated but continue to support SPECIALBUILD and PRIVATEBUILD in favor of
1543    PNG_LIBPNG_BUILD_SPECIAL_STRING and PNG_LIBPNG_BUILD_PRIVATE_STRING.
1544
1545Version 1.2.8beta5 [November 20, 2004]
1546  Use png_ptr->flags instead of png_ptr->transformations to pass
1547    PNG_STRIP_ALPHA info to png_do_strip_filler(), to preserve ABI
1548    compatibility.
1549  Revised handling of SPECIALBUILD, PRIVATEBUILD,
1550    PNG_LIBPNG_BUILD_SPECIAL_STRING and PNG_LIBPNG_BUILD_PRIVATE_STRING.
1551
1552Version 1.2.8rc1 [November 24, 2004]
1553  Moved handling of BUILD macros from pngconf.h to png.h
1554  Added definition of PNG_LIBPNG_BASE_TYPE in png.h, inadvertently
1555    omitted from beta5.
1556  Revised scripts/pngw32.rc
1557  Despammed mailing addresses by masking "@" with "at".
1558  Inadvertently installed a supposedly faster test version of pngrutil.c
1559
1560Version 1.2.8rc2 [November 26, 2004]
1561  Added two missing "\" in png.h
1562  Change tests in pngread.c and pngpread.c to
1563    if (png_ptr->transformations || (png_ptr->flags&PNG_FLAG_STRIP_ALPHA))
1564       png_do_read_transformations(png_ptr);
1565
1566Version 1.2.8rc3 [November 28, 2004]
1567  Reverted pngrutil.c to version libpng-1.2.8beta5.
1568  Added scripts/makefile.elf with supporting code in pngconf.h for symbol
1569    versioning (John Bowler).
1570
1571Version 1.2.8rc4 [November 29, 2004]
1572  Added projects/visualc7 (Simon-pierre).
1573
1574Version 1.2.8rc5 [November 29, 2004]
1575  Fixed new typo in scripts/pngw32.rc
1576
1577Version 1.2.8 [December 3, 2004]
1578  Removed projects/visualc7, added projects/visualc71.
1579
1580Version 1.2.9beta1 [February 21, 2006]
1581  Initialized some structure members in pngwutil.c to avoid gcc-4.0.0 complaints
1582  Revised man page and libpng.txt to make it clear that one should not call
1583    png_read_end or png_write_end after png_read_png or png_write_png.
1584  Updated references to png-mng-implement mailing list.
1585  Fixed an incorrect typecast in pngrutil.c
1586  Added PNG_NO_READ_SUPPORTED conditional for making a write-only library.
1587  Added PNG_NO_WRITE_INTERLACING_SUPPORTED conditional.
1588  Optimized alpha-inversion loops in pngwtran.c
1589  Moved test for nonzero gamma outside of png_build_gamma_table() in pngrtran.c
1590  Make sure num_trans is <= 256 before copying data in png_set_tRNS().
1591  Make sure num_palette is <= 256 before copying data in png_set_PLTE().
1592  Interchanged order of write_swap_alpha and write_invert_alpha transforms.
1593  Added parentheses in the definition of PNG_LIBPNG_BUILD_TYPE (Cosmin).
1594  Optimized zlib window flag (CINFO) in contrib/pngsuite/*.png (Cosmin).
1595  Updated scripts/makefile.bc32 for Borland C++ 5.6 (Cosmin).
1596  Exported png_get_uint_32, png_save_uint_32, png_get_uint_16, png_save_uint_16,
1597    png_get_int_32, png_save_int_32, png_get_uint_31 (Cosmin).
1598  Added type cast (png_byte) in png_write_sCAL() (Cosmin).
1599  Fixed scripts/makefile.cygwin (Christian Biesinger, Cosmin).
1600  Default iTXt support was inadvertently enabled.
1601
1602Version 1.2.9beta2 [February 21, 2006]
1603  Check for png_rgb_to_gray and png_gray_to_rgb read transformations before
1604    checking for png_read_dither in pngrtran.c
1605  Revised checking of chromaticity limits to accommodate extended RGB
1606    colorspace (John Denker).
1607  Changed line endings in some of the project files to CRLF, even in the
1608    "Unix" tar distributions (Cosmin).
1609  Made png_get_int_32 and png_save_int_32 always available (Cosmin).
1610  Updated scripts/pngos2.def, scripts/pngw32.def and projects/wince/png32ce.def
1611    with the newly exported functions.
1612  Eliminated distributions without the "configure" script.
1613  Updated INSTALL instructions.
1614
1615Version 1.2.9beta3 [February 24, 2006]
1616  Fixed CRCRLF line endings in contrib/visupng/VisualPng.dsp
1617  Made libpng.pc respect EXEC_PREFIX (D. P. Kreil, J. Bowler)
1618  Removed reference to pngasmrd.h from Makefile.am
1619  Renamed CHANGES to ChangeLog.
1620  Renamed LICENSE to COPYING.
1621  Renamed ANNOUNCE to NEWS.
1622  Created AUTHORS file.
1623
1624Version 1.2.9beta4 [March 3, 2006]
1625  Changed definition of PKGCONFIG from $prefix/lib to $libdir in configure.ac
1626  Reverted to filenames LICENSE and ANNOUNCE; removed AUTHORS and COPYING.
1627  Removed newline from the end of some error and warning messages.
1628  Removed test for sqrt() from configure.ac and configure.
1629  Made swap tables in pngtrans.c PNG_CONST (Carlo Bramix).
1630  Disabled default iTXt support that was inadvertently enabled in
1631    libpng-1.2.9beta1.
1632  Added "OS2" to list of systems that don't need underscores, in pnggccrd.c
1633  Removed libpng version and date from *.c files.
1634
1635Version 1.2.9beta5 [March 4, 2006]
1636  Removed trailing blanks from source files.
1637  Put version and date of latest change in each source file, and changed
1638    copyright year accordingly.
1639  More cleanup of configure.ac, Makefile.am, and associated scripts.
1640  Restored scripts/makefile.elf which was inadvertently deleted.
1641
1642Version 1.2.9beta6 [March 6, 2006]
1643  Fixed typo (RELEASE) in configuration files.
1644
1645Version 1.2.9beta7 [March 7, 2006]
1646  Removed libpng.vers and libpng.sym from libpng12_la_SOURCES in Makefile.am
1647  Fixed inconsistent #ifdef's around png_sig_bytes() and png_set_sCAL_s()
1648    in png.h.
1649  Updated makefile.elf as suggested by debian.
1650  Made cosmetic changes to some makefiles, adding LN_SF and other macros.
1651  Made some makefiles accept "exec_prefix".
1652
1653Version 1.2.9beta8 [March 9, 2006]
1654  Fixed some "#if defined (..." which should be "#if defined(..."
1655    Bug introduced in libpng-1.2.8.
1656  Fixed inconsistency in definition of png_default_read_data()
1657  Restored blank that was lost from makefile.sggcc "clean" target in beta7.
1658  Revised calculation of "current" and "major" for irix in ltmain.sh
1659  Changed "mkdir" to "MKDIR_P" in some makefiles.
1660  Separated PNG_EXPAND and PNG_EXPAND_tRNS.
1661  Added png_set_expand_gray_1_2_4_to_8() and deprecated
1662    png_set_gray_1_2_4_to_8() which also expands tRNS to alpha.
1663
1664Version 1.2.9beta9 [March 10, 2006]
1665  Include "config.h" in pngconf.h when available.
1666  Added some checks for NULL png_ptr or NULL info_ptr (timeless)
1667
1668Version 1.2.9beta10 [March 20, 2006]
1669  Removed extra CR from contrib/visualpng/VisualPng.dsw (Cosmin)
1670  Made pnggccrd.c PIC-compliant (Christian Aichinger).
1671  Added makefile.mingw (Wolfgang Glas).
1672  Revised pngconf.h MMX checking.
1673
1674Version 1.2.9beta11 [March 22, 2006]
1675  Fixed out-of-order declaration in pngwrite.c that was introduced in beta9
1676  Simplified some makefiles by using LIBSO, LIBSOMAJ, and LIBSOVER macros.
1677
1678Version 1.2.9rc1 [March 31, 2006]
1679  Defined PNG_USER_PRIVATEBUILD when including "pngusr.h" (Cosmin).
1680  Removed nonsensical assertion check from pngtest.c (Cosmin).
1681
1682Version 1.2.9 [April 14, 2006]
1683  Revised makefile.beos and added "none" selector in ltmain.sh
1684
1685Version 1.2.10beta1 [April 15, 2006]
1686  Renamed "config.h" to "png_conf.h" and revised Makefile.am to add
1687    -DPNG_BUILDING_LIBPNG to compile directive, and modified pngconf.h
1688    to include png_conf.h only when PNG_BUILDING_LIBPNG is defined.
1689
1690Version 1.2.10beta2 [April 15, 2006]
1691  Manually updated Makefile.in and configure.  Changed png_conf.h.in
1692    back to config.h.
1693
1694Version 1.2.10beta3 [April 15, 2006]
1695  Change png_conf.h back to config.h in pngconf.h.
1696
1697Version 1.2.10beta4 [April 16, 2006]
1698  Change PNG_BUILDING_LIBPNG to PNG_CONFIGURE_LIBPNG in config/Makefile*.
1699
1700Version 1.2.10beta5 [April 16, 2006]
1701  Added a configure check for compiling assembler code in pnggccrd.c
1702
1703Version 1.2.10beta6 [April 17, 2006]
1704  Revised the configure check for pnggccrd.c
1705  Moved -DPNG_CONFIGURE_LIBPNG into @LIBPNG_DEFINES@
1706  Added @LIBPNG_DEFINES@ to arguments when building libpng.sym
1707
1708Version 1.2.10beta7 [April 18, 2006]
1709  Change "exec_prefix=$prefix" to "exec_prefix=$(prefix)" in makefiles.
1710
1711Version 1.2.10rc1 [April 19, 2006]
1712  Ensure pngconf.h doesn't define both PNG_USE_PNGGCCRD and PNG_USE_PNGVCRD
1713  Fixed "LN_FS" typo in makefile.sco and makefile.solaris.
1714
1715Version 1.2.10rc2 [April 20, 2006]
1716  Added a backslash between -DPNG_CONFIGURE_LIBPNG and -DPNG_NO_ASSEMBLER_CODE
1717   in configure.ac and configure
1718  Made the configure warning about versioned symbols less arrogant.
1719
1720Version 1.2.10rc3 [April 21, 2006]
1721  Added a note in libpng.txt that png_set_sig_bytes(8) can be used when
1722    writing an embedded PNG without the 8-byte signature.
1723  Revised makefiles and configure to avoid making links to libpng.so.*
1724
1725Version 1.2.10 [April 23, 2006]
1726  Reverted configure to "rc2" state.
1727
1728Version 1.2.11beta1 [May 31, 2006]
1729  scripts/libpng.pc.in contained "configure" style version info and would
1730    not work with makefiles.
1731  The shared-library makefiles were linking to libpng.so.0 instead of
1732    libpng.so.3 compatibility as the library.
1733
1734Version 1.2.11beta2 [June 2, 2006]
1735  Increased sprintf buffer from 50 to 52 chars in pngrutil.c to avoid
1736    buffer overflow.
1737  Fixed bug in example.c (png_set_palette_rgb -> png_set_palette_to_rgb)
1738
1739Version 1.2.11beta3 [June 5, 2006]
1740  Prepended "#! /bin/sh" to ltmail.sh and contrib/pngminus/*.sh (Cosmin).
1741  Removed the accidental leftover Makefile.in~ (Cosmin).
1742  Avoided potential buffer overflow and optimized buffer in
1743    png_write_sCAL(), png_write_sCAL_s() (Cosmin).
1744  Removed the include directories and libraries from CFLAGS and LDFLAGS
1745    in scripts/makefile.gcc (Nelson A. de Oliveira, Cosmin).
1746
1747Version 1.2.11beta4 [June 6, 2006]
1748  Allow zero-length IDAT chunks after the entire zlib datastream, but not
1749    after another intervening chunk type.
1750
1751Version 1.0.19rc1, 1.2.11rc1 [June 13, 2006]
1752  Deleted extraneous square brackets from [config.h] in configure.ac
1753
1754Version 1.0.19rc2, 1.2.11rc2 [June 14, 2006]
1755  Added prototypes for PNG_INCH_CONVERSIONS functions to png.h
1756  Revised INSTALL and autogen.sh
1757  Fixed typo in several makefiles (-W1 should be -Wl)
1758  Added typedef for png_int_32 and png_uint_32 on 64-bit systems.
1759
1760Version 1.0.19rc3, 1.2.11rc3 [June 15, 2006]
1761  Removed the new typedefs for 64-bit systems (delay until version 1.4.0)
1762  Added one zero element to png_gamma_shift[] array in pngrtran.c to avoid
1763    reading out of bounds.
1764
1765Version 1.0.19rc4, 1.2.11rc4 [June 15, 2006]
1766  Really removed the new typedefs for 64-bit systems.
1767
1768Version 1.0.19rc5, 1.2.11rc5 [June 22, 2006]
1769  Removed png_sig_bytes entry from scripts/pngw32.def
1770
1771Version 1.0.19, 1.2.11 [June 26, 2006]
1772  None.
1773
1774Version 1.0.20, 1.2.12 [June 27, 2006]
1775  Really increased sprintf buffer from 50 to 52 chars in pngrutil.c to avoid
1776    buffer overflow.
1777
1778Version 1.2.13beta1 [October 2, 2006]
1779  Removed AC_FUNC_MALLOC from configure.ac
1780  Work around Intel-Mac compiler bug by setting PNG_NO_MMX_CODE in pngconf.h
1781  Change "logical" to "bitwise" throughout documentation.
1782  Detect and fix attempt to write wrong iCCP profile length (CVE-2006-7244)
1783
1784Version 1.0.21, 1.2.13 [November 14, 2006]
1785  Fix potential buffer overflow in sPLT chunk handler.
1786  Fix Makefile.am to not try to link to noexistent files.
1787  Check all exported functions for NULL png_ptr.
1788
1789Version 1.2.14beta1 [November 17, 2006]
1790  Relocated three misplaced tests for NULL png_ptr.
1791  Built Makefile.in with automake-1.9.6 instead of 1.9.2.
1792  Build configure with autoconf-2.60 instead of 2.59
1793
1794Version 1.2.14beta2 [November 17, 2006]
1795  Added some typecasts in png_zalloc().
1796
1797Version 1.2.14rc1 [November 20, 2006]
1798  Changed "strtod" to "png_strtod" in pngrutil.c
1799
1800Version 1.0.22, 1.2.14    [November 27, 2006]
1801  Added missing "$(srcdir)" in Makefile.am and Makefile.in
1802
1803Version 1.2.15beta1 [December 3, 2006]
1804  Generated configure with autoconf-2.61 instead of 2.60
1805  Revised configure.ac to update libpng.pc and libpng-config.
1806
1807Version 1.2.15beta2 [December 3, 2006]
1808  Always export MMX asm functions, just stubs if not building pnggccrd.c
1809
1810Version 1.2.15beta3 [December 4, 2006]
1811  Add "png_bytep" typecast to profile while calculating length in pngwutil.c
1812
1813Version 1.2.15beta4 [December 7, 2006]
1814  Added scripts/CMakeLists.txt
1815  Changed PNG_NO_ASSEMBLER_CODE to PNG_NO_MMX_CODE in scripts, like 1.4.0beta
1816
1817Version 1.2.15beta5 [December 7, 2006]
1818  Changed some instances of PNG_ASSEMBLER_* to PNG_MMX_* in pnggccrd.c
1819  Revised scripts/CMakeLists.txt
1820
1821Version 1.2.15beta6 [December 13, 2006]
1822  Revised scripts/CMakeLists.txt and configure.ac
1823
1824Version 1.2.15rc1 [December 18, 2006]
1825  Revised scripts/CMakeLists.txt
1826
1827Version 1.2.15rc2 [December 21, 2006]
1828  Added conditional #undef jmpbuf in pngtest.c to undo #define in AIX headers.
1829  Added scripts/makefile.nommx
1830
1831Version 1.2.15rc3 [December 25, 2006]
1832  Fixed shared library numbering error that was introduced in 1.2.15beta6.
1833
1834Version 1.2.15rc4 [December 27, 2006]
1835  Fixed handling of rgb_to_gray when png_ptr->color.gray isn't set.
1836
1837Version 1.2.15rc5 [December 31, 2006]
1838  Revised handling of rgb_to_gray.
1839
1840Version 1.2.15 [January 5, 2007]
1841  Added some (unsigned long) typecasts in pngtest.c to avoid printing errors.
1842
1843Version 1.2.16beta1 [January 6, 2007]
1844  Fix bugs in makefile.nommx
1845
1846Version 1.2.16beta2 [January 16, 2007]
1847  Revised scripts/CMakeLists.txt
1848
1849Version 1.2.16 [January 31, 2007]
1850  No changes.
1851
1852Version 1.2.17beta1 [March 6, 2007]
1853  Revised scripts/CMakeLists.txt to install both shared and static libraries.
1854  Deleted a redundant line from pngset.c.
1855
1856Version 1.2.17beta2 [April 26, 2007]
1857  Relocated misplaced test for png_ptr == NULL in pngpread.c
1858  Change "==" to "&" for testing PNG_RGB_TO_GRAY_ERR & PNG_RGB_TO_GRAY_WARN
1859    flags.
1860  Changed remaining instances of PNG_ASSEMBLER_* to PNG_MMX_*
1861  Added pngerror() when write_IHDR fails in deflateInit2().
1862  Added "const" to some array declarations.
1863  Mention examples of libpng usage in the libpng*.txt and libpng.3 documents.
1864
1865Version 1.2.17rc1 [May 4, 2007]
1866  No changes.
1867
1868Version 1.2.17rc2 [May 8, 2007]
1869  Moved several PNG_HAVE_* macros out of PNG_INTERNAL because applications
1870    calling set_unknown_chunk_location() need them.
1871  Changed transformation flag from PNG_EXPAND_tRNS to PNG_EXPAND in
1872    png_set_expand_gray_1_2_4_to_8().
1873  Added png_ptr->unknown_chunk to hold working unknown chunk data, so it
1874    can be free'ed in case of error.  Revised unknown chunk handling in
1875    pngrutil.c and pngpread.c to use this structure.
1876
1877Version 1.2.17rc3 [May 8, 2007]
1878  Revised symbol-handling in configure script.
1879
1880Version 1.2.17rc4 [May 10, 2007]
1881  Revised unknown chunk handling to avoid storing unknown critical chunks.
1882
1883Version 1.0.25 [May 15, 2007]
1884Version 1.2.17 [May 15, 2007]
1885  Added "png_ptr->num_trans=0" before error return in png_handle_tRNS,
1886    to eliminate a vulnerability (CVE-2007-2445, CERT VU#684664)
1887
1888Version 1.0.26 [May 15, 2007]
1889Version 1.2.18 [May 15, 2007]
1890  Reverted the libpng-1.2.17rc3 change to symbol-handling in configure script
1891
1892Version 1.2.19beta1 [May 18, 2007]
1893  Changed "const static" to "static PNG_CONST" everywhere, mostly undoing
1894    change of libpng-1.2.17beta2.  Changed other "const" to "PNG_CONST"
1895  Changed some handling of unused parameters, to avoid compiler warnings.
1896    "if (unused == NULL) return;" becomes "unused = unused".
1897
1898Version 1.2.19beta2 [May 18, 2007]
1899  Only use the valid bits of tRNS value in png_do_expand() (Brian Cartier)
1900
1901Version 1.2.19beta3 [May 19, 2007]
1902  Add some "png_byte" typecasts in png_check_keyword() and write new_key
1903  instead of key in zTXt chunk (Kevin Ryde).
1904
1905Version 1.2.19beta4 [May 21, 2007]
1906  Add png_snprintf() function and use it in place of sprint() for improved
1907    defense against buffer overflows.
1908
1909Version 1.2.19beta5 [May 21, 2007]
1910  Fixed png_handle_tRNS() to only use the valid bits of tRNS value.
1911  Changed handling of more unused parameters, to avoid compiler warnings.
1912  Removed some PNG_CONST in pngwutil.c to avoid compiler warnings.
1913
1914Version 1.2.19beta6 [May 22, 2007]
1915  Added some #ifdef PNG_MMX_CODE_SUPPORTED where needed in pngvcrd.c
1916  Added a special "_MSC_VER" case that defines png_snprintf to _snprintf
1917
1918Version 1.2.19beta7 [May 22, 2007]
1919  Squelched png_squelch_warnings() in pnggccrd.c and added
1920    an #ifdef PNG_MMX_CODE_SUPPORTED block around the declarations that caused
1921    the warnings that png_squelch_warnings was squelching.
1922
1923Version 1.2.19beta8 [May 22, 2007]
1924  Removed __MMX__ from test in pngconf.h.
1925
1926Version 1.2.19beta9 [May 23, 2007]
1927  Made png_squelch_warnings() available via PNG_SQUELCH_WARNINGS macro.
1928  Revised png_squelch_warnings() so it might work.
1929  Updated makefile.sgcc and makefile.solaris; added makefile.solaris-x86.
1930
1931Version 1.2.19beta10 [May 24, 2007]
1932  Resquelched png_squelch_warnings(), use "__attribute__((used))" instead.
1933
1934Version 1.4.0beta1 [April 20, 2006]
1935  Enabled iTXt support (changes png_struct, thus requires so-number change).
1936  Cleaned up PNG_ASSEMBLER_CODE_SUPPORTED vs PNG_MMX_CODE_SUPPORTED
1937  Eliminated PNG_1_0_X and PNG_1_2_X macros.
1938  Removed deprecated functions png_read_init, png_write_init, png_info_init,
1939    png_permit_empty_plte, png_set_gray_1_2_4_to_8, png_check_sig, and
1940    removed the deprecated macro PNG_MAX_UINT.
1941  Moved "PNG_INTERNAL" parts of png.h and pngconf.h into pngintrn.h
1942  Removed many WIN32_WCE #ifdefs (Cosmin).
1943  Reduced dependency on C-runtime library when on Windows (Simon-Pierre)
1944  Replaced sprintf() with png_sprintf() (Simon-Pierre)
1945
1946Version 1.4.0beta2 [April 20, 2006]
1947  Revised makefiles and configure to avoid making links to libpng.so.*
1948  Moved some leftover MMX-related defines from pngconf.h to pngintrn.h
1949  Updated scripts/pngos2.def, pngw32.def, and projects/wince/png32ce.def
1950
1951Version 1.4.0beta3 [May 10, 2006]
1952  Updated scripts/pngw32.def to comment out MMX functions.
1953  Added PNG_NO_GET_INT_32 and PNG_NO_SAVE_INT_32 macros.
1954  Scripts/libpng.pc.in contained "configure" style version info and would
1955    not work with makefiles.
1956  Revised pngconf.h and added pngconf.h.in, so makefiles and configure can
1957    pass defines to libpng and applications.
1958
1959Version 1.4.0beta4 [May 11, 2006]
1960  Revised configure.ac, Makefile.am, and many of the makefiles to write
1961    their defines in pngconf.h.
1962
1963Version 1.4.0beta5 [May 15, 2006]
1964  Added a missing semicolon in Makefile.am and Makefile.in
1965  Deleted extraneous square brackets from configure.ac
1966
1967Version 1.4.0beta6 [June 2, 2006]
1968  Increased sprintf buffer from 50 to 52 chars in pngrutil.c to avoid
1969    buffer overflow.
1970  Changed sonum from 0 to 1.
1971  Removed unused prototype for png_check_sig() from png.h
1972
1973Version 1.4.0beta7 [June 16, 2006]
1974  Exported png_write_sig (Cosmin).
1975  Optimized buffer in png_handle_cHRM() (Cosmin).
1976  Set pHYs = 2835 x 2835 pixels per meter, and added
1977    sCAL = 0.352778e-3 x 0.352778e-3 meters, in pngtest.png (Cosmin).
1978  Added png_set_benign_errors(), png_benign_error(), png_chunk_benign_error().
1979  Added typedef for png_int_32 and png_uint_32 on 64-bit systems.
1980  Added "(unsigned long)" typecast on png_uint_32 variables in printf lists.
1981
1982Version 1.4.0beta8 [June 22, 2006]
1983  Added demonstration of user chunk support in pngtest.c, to support the
1984    public sTER chunk and a private vpAg chunk.
1985
1986Version 1.4.0beta9 [July 3, 2006]
1987  Removed ordinals from scripts/pngw32.def and removed png_info_int and
1988    png_set_gray_1_2_4_to_8 entries.
1989  Inline call of png_get_uint_32() in png_get_uint_31().
1990  Use png_get_uint_31() to get vpAg width and height in pngtest.c
1991  Removed WINCE and Netware projects.
1992  Removed standalone Y2KINFO file.
1993
1994Version 1.4.0beta10 [July 12, 2006]
1995  Eliminated automatic copy of pngconf.h to pngconf.h.in from configure and
1996    some makefiles, because it was not working reliably.  Instead, distribute
1997    pngconf.h.in along with pngconf.h and cause configure and some of the
1998    makefiles to update pngconf.h from pngconf.h.in.
1999  Added pngconf.h to DEPENDENCIES in Makefile.am
2000
2001Version 1.4.0beta11 [August 19, 2006]
2002  Removed AC_FUNC_MALLOC from configure.ac.
2003  Added a warning when writing iCCP profile with mismatched profile length.
2004  Patched pnggccrd.c to assemble on x86_64 platforms.
2005  Moved chunk header reading into a separate function png_read_chunk_header()
2006    in pngrutil.c.  The chunk header (len+sig) is now serialized in a single
2007    operation (Cosmin).
2008  Implemented support for I/O states. Added png_ptr member io_state, and
2009    functions png_get_io_chunk_name() and png_get_io_state() in pngget.c
2010    (Cosmin).
2011  Added png_get_io_chunk_name and png_get_io_state to scripts/*.def (Cosmin).
2012  Renamed scripts/pngw32.* to scripts/pngwin.* (Cosmin).
2013  Removed the include directories and libraries from CFLAGS and LDFLAGS
2014    in scripts/makefile.gcc (Cosmin).
2015  Used png_save_uint_32() to set vpAg width and height in pngtest.c (Cosmin).
2016  Cast to proper type when getting/setting vpAg units in pngtest.c (Cosmin).
2017  Added pngintrn.h to the Visual C++ projects (Cosmin).
2018  Removed scripts/list (Cosmin).
2019  Updated copyright year in scripts/pngwin.def (Cosmin).
2020  Removed PNG_TYPECAST_NULL and used standard NULL consistently (Cosmin).
2021  Disallowed the user to redefine png_size_t, and enforced a consistent use
2022    of png_size_t across libpng (Cosmin).
2023  Changed the type of png_ptr->rowbytes, PNG_ROWBYTES() and friends
2024    to png_size_t (Cosmin).
2025  Removed png_convert_size() and replaced png_sizeof with sizeof (Cosmin).
2026  Removed some unnecessary type casts (Cosmin).
2027  Changed prototype of png_get_compression_buffer_size() and
2028    png_set_compression_buffer_size() to work with png_size_t instead of
2029    png_uint_32 (Cosmin).
2030  Removed png_memcpy_check() and png_memset_check() (Cosmin).
2031  Fixed a typo (png_byte --> png_bytep) in libpng.3 and libpng.txt (Cosmin).
2032  Clarified that png_zalloc() does not clear the allocated memory,
2033    and png_zalloc() and png_zfree() cannot be PNGAPI (Cosmin).
2034  Renamed png_mem_size_t to png_alloc_size_t, fixed its definition in
2035    pngconf.h, and used it in all memory allocation functions (Cosmin).
2036  Renamed pngintrn.h to pngpriv.h, added a comment at the top of the file
2037    mentioning that the symbols declared in that file are private, and
2038    updated the scripts and the Visual C++ projects accordingly (Cosmin).
2039  Removed circular references between pngconf.h and pngconf.h.in in
2040    scripts/makefile.vc*win32 (Cosmin).
2041  Removing trailing '.' from the warning and error messages (Cosmin).
2042  Added pngdefs.h that is built by makefile or configure, instead of
2043    pngconf.h.in (Glenn).
2044  Detect and fix attempt to write wrong iCCP profile length.
2045
2046Version 1.4.0beta12 [October 19, 2006]
2047  Changed "logical" to "bitwise" in the documentation.
2048  Work around Intel-Mac compiler bug by setting PNG_NO_MMX_CODE in pngconf.h
2049  Add a typecast to stifle compiler warning in pngrutil.c
2050
2051Version 1.4.0beta13 [November 10, 2006]
2052  Fix potential buffer overflow in sPLT chunk handler.
2053  Fix Makefile.am to not try to link to noexistent files.
2054
2055Version 1.4.0beta14 [November 15, 2006]
2056  Check all exported functions for NULL png_ptr.
2057
2058Version 1.4.0beta15 [November 17, 2006]
2059  Relocated two misplaced tests for NULL png_ptr.
2060  Built Makefile.in with automake-1.9.6 instead of 1.9.2.
2061  Build configure with autoconf-2.60 instead of 2.59
2062  Add "install: all" in Makefile.am so "configure; make install" will work.
2063
2064Version 1.4.0beta16 [November 17, 2006]
2065  Added a typecast in png_zalloc().
2066
2067Version 1.4.0beta17 [December 4, 2006]
2068  Changed "new_key[79] = '\0';" to "(*new_key)[79] = '\0';" in pngwutil.c
2069  Add "png_bytep" typecast to profile while calculating length in pngwutil.c
2070
2071Version 1.4.0beta18 [December 7, 2006]
2072  Added scripts/CMakeLists.txt
2073
2074Version 1.4.0beta19 [May 16, 2007]
2075  Revised scripts/CMakeLists.txt
2076  Rebuilt configure and Makefile.in with newer tools.
2077  Added conditional #undef jmpbuf in pngtest.c to undo #define in AIX headers.
2078  Added scripts/makefile.nommx
2079
2080Version 1.4.0beta20 [July 9, 2008]
2081  Moved several PNG_HAVE_* macros from pngpriv.h to png.h because applications
2082    calling set_unknown_chunk_location() need them.
2083  Moved several macro definitions from pngpriv.h to pngconf.h
2084  Merge with changes to the 1.2.X branch, as of 1.2.30beta04.
2085  Deleted all use of the MMX assembler code and Intel-licensed optimizations.
2086  Revised makefile.mingw
2087
2088Version 1.4.0beta21 [July 21, 2008]
2089  Moved local array "chunkdata" from pngrutil.c to the png_struct, so
2090    it will be freed by png_read_destroy() in case of a read error (Kurt
2091    Christensen).
2092
2093Version 1.4.0beta22 [July 21, 2008]
2094  Change "purpose" and "buffer" to png_ptr->chunkdata to avoid memory leaking.
2095
2096Version 1.4.0beta23 [July 22, 2008]
2097  Change "chunkdata = NULL" to "png_ptr->chunkdata = NULL" several places in
2098    png_decompress_chunk().
2099
2100Version 1.4.0beta24 [July 25, 2008]
2101  Change all remaining "chunkdata" to "png_ptr->chunkdata" in
2102    png_decompress_chunk(), and remove "chunkdata" from parameter list.
2103  Put a call to png_check_chunk_name() in png_read_chunk_header().
2104  Revised png_check_chunk_name() to reject a name with a lowercase 3rd byte.
2105  Removed two calls to png_check_chunk_name() occuring later in the process.
2106  Define PNG_NO_ERROR_NUMBERS by default in pngconf.h
2107
2108Version 1.4.0beta25 [July 30, 2008]
2109  Added a call to png_check_chunk_name() in pngpread.c
2110  Reverted png_check_chunk_name() to accept a name with a lowercase 3rd byte.
2111  Added png_push_have_buffer() function to pngpread.c
2112  Eliminated PNG_BIG_ENDIAN_SUPPORTED and associated png_get_* macros.
2113  Made inline expansion of png_get_*() optional with PNG_USE_READ_MACROS.
2114  Eliminated all PNG_USELESS_TESTS and PNG_CORRECT_PALETTE_SUPPORTED code.
2115  Synced contrib directory and configure files with libpng-1.2.30beta06.
2116  Eliminated no-longer-used pngdefs.h (but it's still built in the makefiles)
2117  Relocated a misplaced "#endif /* PNG_NO_WRITE_FILTER */" in pngwutil.c
2118
2119Version 1.4.0beta26 [August 4, 2008]
2120  Removed png_push_have_buffer() function in pngpread.c.  It increased the
2121    compiled library size slightly.
2122  Changed "-Wall" to "-W -Wall" in the CFLAGS in all makefiles (Cosmin Truta)
2123  Declared png_ptr "volatile" in pngread.c and pngwrite.c to avoid warnings.
2124  Updated contrib/visupng/cexcept.h to version 2.0.1
2125  Added PNG_LITERAL_CHARACTER macros for #, [, and ].
2126
2127Version 1.4.0beta27 [August 5, 2008]
2128  Revised usage of PNG_LITERAL_SHARP in pngerror.c.
2129  Moved newline character from individual png_debug messages into the
2130    png_debug macros.
2131  Allow user to #define their own png_debug, png_debug1, and png_debug2.
2132
2133Version 1.4.0beta28 [August 5, 2008]
2134  Revised usage of PNG_LITERAL_SHARP in pngerror.c.
2135  Added PNG_STRING_NEWLINE macro
2136
2137Version 1.4.0beta29 [August 9, 2008]
2138  Revised usage of PNG_STRING_NEWLINE to work on non-ISO compilers.
2139  Added PNG_STRING_COPYRIGHT macro.
2140  Added non-ISO versions of png_debug macros.
2141
2142Version 1.4.0beta30 [August 14, 2008]
2143  Added premultiplied alpha feature (Volker Wiendl).
2144
2145Version 1.4.0beta31 [August 18, 2008]
2146  Moved png_set_premultiply_alpha from pngtrans.c to pngrtran.c
2147  Removed extra crc check at the end of png_handle_cHRM().  Bug introduced
2148    in libpng-1.4.0beta20.
2149
2150Version 1.4.0beta32 [August 19, 2008]
2151  Added PNG_WRITE_FLUSH_SUPPORTED block around new png_flush() call.
2152  Revised PNG_NO_STDIO version of png_write_flush()
2153
2154Version 1.4.0beta33 [August 20, 2008]
2155  Added png_get|set_chunk_cache_max() to limit the total number of sPLT,
2156    text, and unknown chunks that can be stored.
2157
2158Version 1.4.0beta34 [September 6, 2008]
2159  Shortened tIME_string to 29 bytes in pngtest.c
2160  Fixed off-by-one error introduced in png_push_read_zTXt() function in
2161    libpng-1.2.30beta04/pngpread.c (Harald van Dijk)
2162
2163Version 1.4.0beta35 [October 6, 2008]
2164  Changed "trans_values" to "trans_color".
2165  Changed so-number from 0 to 14.  Some OS do not like 0.
2166  Revised makefile.darwin to fix shared library numbering.
2167  Change png_set_gray_1_2_4_to_8() to png_set_expand_gray_1_2_4_to_8()
2168    in example.c (debian bug report)
2169
2170Version 1.4.0beta36 [October 25, 2008]
2171  Sync with tEXt vulnerability fix in libpng-1.2.33rc02.
2172
2173Version 1.4.0beta37 [November 13, 2008]
2174  Added png_check_cHRM in png.c and moved checking from pngget.c, pngrutil.c,
2175    and pngwrite.c
2176
2177Version 1.4.0beta38 [November 22, 2008]
2178  Added check for zero-area RGB cHRM triangle in png_check_cHRM() and
2179    png_check_cHRM_fixed().
2180
2181Version 1.4.0beta39 [November 23, 2008]
2182  Revised png_warning() to write its message on standard output by default
2183    when warning_fn is NULL.
2184
2185Version 1.4.0beta40 [November 24, 2008]
2186  Eliminated png_check_cHRM().  Instead, always use png_check_cHRM_fixed().
2187  In png_check_cHRM_fixed(), ensure white_y is > 0, and removed redundant
2188    check for all-zero coordinates that is detected by the triangle check.
2189
2190Version 1.4.0beta41 [November 26, 2008]
2191  Fixed string vs pointer-to-string error in png_check_keyword().
2192  Rearranged test expressions in png_check_cHRM_fixed() to avoid internal
2193    overflows.
2194  Added PNG_NO_CHECK_cHRM conditional.
2195
2196Version 1.4.0beta42, 43 [December 1, 2008]
2197  Merge png_debug with version 1.2.34beta04.
2198
2199Version 1.4.0beta44 [December 6, 2008]
2200  Removed redundant check for key==NULL before calling png_check_keyword()
2201    to ensure that new_key gets initialized and removed extra warning
2202    (Merge with version 1.2.34beta05 -- Arvan Pritchard).
2203
2204Version 1.4.0beta45 [December 9, 2008]
2205  In png_write_png(), respect the placement of the filler bytes in an earlier
2206    call to png_set_filler() (Jim Barry).
2207
2208Version 1.4.0beta46 [December 10, 2008]
2209  Undid previous change and added PNG_TRANSFORM_STRIP_FILLER_BEFORE and
2210    PNG_TRANSFORM_STRIP_FILLER_AFTER conditionals and deprecated
2211    PNG_TRANSFORM_STRIP_FILLER (Jim Barry).
2212
2213Version 1.4.0beta47 [December 15, 2008]
2214  Support for dithering was disabled by default, because it has never
2215    been well tested and doesn't work very well.  The code has not
2216    been removed, however, and can be enabled by building libpng with
2217    PNG_READ_DITHER_SUPPORTED defined.
2218
2219Version 1.4.0beta48 [February 14, 2009]
2220  Added new exported function png_calloc().
2221  Combined several instances of png_malloc(); png_memset() into png_calloc().
2222  Removed prototype for png_freeptr() that was added in libpng-1.4.0beta24
2223    but was never defined.
2224
2225Version 1.4.0beta49 [February 28, 2009]
2226  Added png_fileno() macro to pngconf.h, used in pngwio.c
2227  Corrected order of #ifdef's in png_debug definition in png.h
2228  Fixed bug introduced in libpng-1.4.0beta48 with the memset arguments
2229    for pcal_params.
2230  Fixed order of #ifdef directives in the png_debug defines in png.h
2231    (bug introduced in libpng-1.2.34/1.4.0beta29).
2232  Revised comments in png_set_read_fn() and png_set_write_fn().
2233
2234Version 1.4.0beta50 [March 18, 2009]
2235  Use png_calloc() instead of png_malloc() to allocate big_row_buf when
2236    reading an interlaced file, to avoid a possible UMR.
2237  Undid revision of PNG_NO_STDIO version of png_write_flush().  Users
2238    having trouble with fflush() can build with PNG_NO_WRITE_FLUSH defined
2239    or supply their own flush_fn() replacement.
2240  Revised libpng*.txt and png.h documentation about use of png_write_flush()
2241    and png_set_write_fn().
2242  Removed fflush() from pngtest.c.
2243  Added "#define PNG_NO_WRITE_FLUSH" to contrib/pngminim/encoder/pngusr.h
2244
2245Version 1.4.0beta51 [March 21, 2009]
2246  Removed new png_fileno() macro from pngconf.h .
2247
2248Version 1.4.0beta52 [March 27, 2009]
2249  Relocated png_do_chop() ahead of building gamma tables in pngrtran.c
2250    This avoids building 16-bit gamma tables unnecessarily.
2251  Removed fflush() from pngtest.c.
2252  Added "#define PNG_NO_WRITE_FLUSH" to contrib/pngminim/encoder/pngusr.h
2253  Added a section on differences between 1.0.x and 1.2.x to libpng.3/libpng.txt
2254
2255Version 1.4.0beta53 [April 1, 2009]
2256  Removed some remaining MMX macros from pngpriv.h
2257  Fixed potential memory leak of "new_name" in png_write_iCCP() (Ralph Giles)
2258
2259Version 1.4.0beta54 [April 13, 2009]
2260  Added "ifndef PNG_SKIP_SETJMP_CHECK" block in pngconf.h to allow
2261    application code writers to bypass the check for multiple inclusion
2262    of setjmp.h when they know that it is safe to ignore the situation.
2263  Eliminated internal use of setjmp() in pngread.c and pngwrite.c
2264  Reordered ancillary chunks in pngtest.png to be the same as what
2265    pngtest now produces, and made some cosmetic changes to pngtest output.
2266  Eliminated deprecated png_read_init_3() and png_write_init_3() functions.
2267
2268Version 1.4.0beta55 [April 15, 2009]
2269  Simplified error handling in pngread.c and pngwrite.c by putting
2270    the new png_read_cleanup() and png_write_cleanup() functions inline.
2271
2272Version 1.4.0beta56 [April 25, 2009]
2273  Renamed "user_chunk_data" to "my_user_chunk_data" in pngtest.c to suppress
2274    "shadowed declaration" warning from gcc-4.3.3.
2275  Renamed "gamma" to "png_gamma" in pngset.c to avoid "shadowed declaration"
2276    warning about a global "gamma" variable in math.h on some platforms.
2277
2278Version 1.4.0beta57 [May 2, 2009]
2279  Removed prototype for png_freeptr() that was added in libpng-1.4.0beta24
2280    but was never defined (again).
2281  Rebuilt configure scripts with autoconf-2.63 instead of 2.62
2282  Removed pngprefs.h and MMX from makefiles
2283
2284Version 1.4.0beta58 [May 14, 2009]
2285  Changed pngw32.def to pngwin.def in makefile.mingw (typo was introduced
2286    in beta57).
2287  Clarified usage of sig_bit versus sig_bit_p in example.c (Vincent Torri)
2288
2289Version 1.4.0beta59 [May 15, 2009]
2290  Reformated sources in libpng style (3-space intentation, comment format)
2291  Fixed typo in libpng docs (PNG_FILTER_AVE should be PNG_FILTER_AVG)
2292  Added sections about the git repository and our coding style to the
2293    documentation
2294  Relocated misplaced #endif in pngwrite.c, sCAL chunk handler.
2295
2296Version 1.4.0beta60 [May 19, 2009]
2297  Conditionally compile png_read_finish_row() which is not used by
2298    progressive readers.
2299  Added contrib/pngminim/preader to demonstrate building minimal progressive
2300    decoder, based on contrib/gregbook with embedded libpng and zlib.
2301
2302Version 1.4.0beta61 [May 20, 2009]
2303  In contrib/pngminim/*, renamed "makefile.std" to "makefile", since there
2304    is only one makefile in those directories, and revised the README files
2305    accordingly.
2306  More reformatting of comments, mostly to capitalize sentences.
2307
2308Version 1.4.0beta62 [June 2, 2009]
2309  Added "#define PNG_NO_WRITE_SWAP" to contrib/pngminim/encoder/pngusr.h
2310    and "define PNG_NO_READ_SWAP" to decoder/pngusr.h and preader/pngusr.h
2311  Reformatted several remaining "else statement" into two lines.
2312  Added a section to the libpng documentation about using png_get_io_ptr()
2313    in configure scripts to detect the presence of libpng.
2314
2315Version 1.4.0beta63 [June 15, 2009]
2316  Revised libpng*.txt and libpng.3 to mention calling png_set_IHDR()
2317    multiple times and to specify the sample order in the tRNS chunk,
2318    because the ISO PNG specification has a typo in the tRNS table.
2319  Changed several PNG_UNKNOWN_CHUNK_SUPPORTED to
2320    PNG_HANDLE_AS_UNKNOWN_SUPPORTED, to make the png_set_keep mechanism
2321    available for ignoring known chunks even when not saving unknown chunks.
2322  Adopted preference for consistent use of "#ifdef" and "#ifndef" versus
2323    "#if defined()" and "if !defined()" where possible.
2324
2325Version 1.4.0beta64 [June 24, 2009]
2326  Eliminated PNG_LEGACY_SUPPORTED code.
2327  Moved the various unknown chunk macro definitions outside of the
2328     PNG_READ|WRITE_ANCILLARY_CHUNK_SUPPORTED blocks.
2329
2330Version 1.4.0beta65 [June 26, 2009]
2331  Added a reference to the libpng license in each file.
2332
2333Version 1.4.0beta66 [June 27, 2009]
2334  Refer to the libpng license instead of the libpng license in each file.
2335
2336Version 1.4.0beta67 [July 6, 2009]
2337  Relocated INVERT_ALPHA within png_read_png() and png_write_png().
2338  Added high-level API transform PNG_TRANSFORM_GRAY_TO_RGB.
2339  Added an "xcode" project to the projects directory (Alam Arias).
2340
2341Version 1.4.0beta68 [July 19, 2009]
2342  Avoid some tests in filter selection in pngwutil.c
2343
2344Version 1.4.0beta69 [July 25, 2009]
2345  Simplified the new filter-selection test.  This runs faster in the
2346    common "PNG_ALL_FILTERS" and PNG_FILTER_NONE cases.
2347  Removed extraneous declaration from the new call to png_read_gray_to_rgb()
2348    (bug introduced in libpng-1.4.0beta67).
2349  Fixed up xcode project (Alam Arias)
2350  Added a prototype for png_64bit_product() in png.c
2351
2352Version 1.4.0beta70 [July 27, 2009]
2353  Avoid a possible NULL dereference in debug build, in png_set_text_2().
2354    (bug introduced in libpng-0.95, discovered by Evan Rouault)
2355
2356Version 1.4.0beta71 [July 29, 2009]
2357  Rebuilt configure scripts with autoconf-2.64.
2358
2359Version 1.4.0beta72 [August 1, 2009]
2360  Replaced *.tar.lzma with *.tar.xz in distribution.  Get the xz codec
2361    from <http://tukaani.org/xz>.
2362
2363Version 1.4.0beta73 [August 1, 2009]
2364  Reject attempt to write iCCP chunk with negative embedded profile length
2365    (JD Chen) (CVE-2009-5063).
2366
2367Version 1.4.0beta74 [August 8, 2009]
2368  Changed png_ptr and info_ptr member "trans" to "trans_alpha".
2369
2370Version 1.4.0beta75 [August 21, 2009]
2371  Removed an extra png_debug() recently added to png_write_find_filter().
2372  Fixed incorrect #ifdef in pngset.c regarding unknown chunk support.
2373
2374Version 1.4.0beta76 [August 22, 2009]
2375  Moved an incorrectly located test in png_read_row() in pngread.c
2376
2377Version 1.4.0beta77 [August 27, 2009]
2378  Removed lpXYZ.tar.bz2 (with CRLF), KNOWNBUG, libpng-x.y.z-KNOWNBUG.txt,
2379    and the "noconfig" files from the distribution.
2380  Moved CMakeLists.txt from scripts into the main libpng directory.
2381  Various bugfixes and improvements to CMakeLists.txt (Philip Lowman)
2382
2383Version 1.4.0beta78 [August 31, 2009]
2384  Converted all PNG_NO_* tests to PNG_*_SUPPORTED everywhere except pngconf.h
2385  Eliminated PNG_NO_FREE_ME and PNG_FREE_ME_SUPPORTED macros.
2386  Use png_malloc plus a loop instead of png_calloc() to initialize
2387    row_pointers in png_read_png().
2388
2389Version 1.4.0beta79 [September 1, 2009]
2390  Eliminated PNG_GLOBAL_ARRAYS and PNG_LOCAL_ARRAYS; always use local arrays.
2391  Eliminated PNG_CALLOC_SUPPORTED macro and always provide png_calloc().
2392
2393Version 1.4.0beta80 [September 17, 2009]
2394  Removed scripts/libpng.icc
2395  Changed typecast of filler from png_byte to png_uint_16 in png_set_filler().
2396    (Dennis Gustafsson)
2397  Fixed typo introduced in beta78 in pngtest.c ("#if def " should be "#ifdef ")
2398
2399Version 1.4.0beta81 [September 23, 2009]
2400  Eliminated unused PNG_FLAG_FREE_* defines from pngpriv.h
2401  Expanded TAB characters in pngrtran.c
2402  Removed PNG_CONST from all "PNG_CONST PNG_CHNK" declarations to avoid
2403    compiler complaints about doubly declaring things "const".
2404  Changed all "#if [!]defined(X)" to "if[n]def X" where possible.
2405  Eliminated unused png_ptr->row_buf_size
2406
2407Version 1.4.0beta82 [September 25, 2009]
2408  Moved redundant IHDR checking into new png_check_IHDR() in png.c
2409    and report all errors found in the IHDR data.
2410  Eliminated useless call to png_check_cHRM() from pngset.c
2411
2412Version 1.4.0beta83 [September 25, 2009]
2413  Revised png_check_IHDR() to eliminate bogus complaint about filter_type.
2414
2415Version 1.4.0beta84 [September 30, 2009]
2416  Fixed some inconsistent indentation in pngconf.h
2417  Revised png_check_IHDR() to add a test for width variable less than 32-bit.
2418
2419Version 1.4.0beta85 [October 1, 2009]
2420  Revised png_check_IHDR() again, to check info_ptr members instead of
2421    the contents of the returned parameters.
2422
2423Version 1.4.0beta86 [October 9, 2009]
2424  Updated the "xcode" project (Alam Arias).
2425  Eliminated a shadowed declaration of "pp" in png_handle_sPLT().
2426
2427Version 1.4.0rc01 [October 19, 2009]
2428  Trivial cosmetic changes.
2429
2430Version 1.4.0beta87 [October 30, 2009]
2431  Moved version 1.4.0 back into beta.
2432
2433Version 1.4.0beta88 [October 30, 2009]
2434  Revised libpng*.txt section about differences between 1.2.x and 1.4.0
2435    because most of the new features have now been ported back to 1.2.41
2436
2437Version 1.4.0beta89 [November 1, 2009]
2438  More bugfixes and improvements to CMakeLists.txt (Philip Lowman)
2439  Removed a harmless extra png_set_invert_alpha() from pngwrite.c
2440  Apply png_user_chunk_cache_max within png_decompress_chunk().
2441  Merged libpng-1.2.41.txt with libpng-1.4.0.txt where appropriate.
2442
2443Version 1.4.0beta90 [November 2, 2009]
2444  Removed all remaining WIN32_WCE #ifdefs except those involving the
2445    time.h "tm" structure
2446
2447Version 1.4.0beta91 [November 3, 2009]
2448  Updated scripts/pngw32.def and projects/wince/png32ce.def
2449  Copied projects/wince/png32ce.def to the scripts directory.
2450  Added scripts/makefile.wce
2451  Patched ltmain.sh for wince support.
2452  Added PNG_CONVERT_tIME_SUPPORTED macro.
2453
2454Version 1.4.0beta92 [November 4, 2009]
2455  Make inclusion of time.h in pngconf.h depend on PNG_CONVERT_tIME_SUPPORTED
2456  Make #define PNG_CONVERT_tIME_SUPPORTED depend on PNG_WRITE_tIME_SUPPORTED
2457  Revised libpng*.txt to describe differences from 1.2.40 to 1.4.0 (instead
2458    of differences from 1.2.41 to 1.4.0)
2459
2460Version 1.4.0beta93 [November 7, 2009]
2461  Added PNG_DEPSTRUCT, PNG_DEPRECATED, PNG_USE_RESULT, PNG_NORETURN, and
2462    PNG_ALLOCATED macros to detect deprecated direct access to the
2463    png_struct or info_struct members and other deprecated usage in
2464    applications (John Bowler).
2465  Updated scripts/makefile* to add "-DPNG_CONFIGURE_LIBPNG" to CFLAGS,
2466    to prevent warnings about direct access to png structs by libpng
2467    functions while building libpng.  They need to be tested, especially
2468    those using compilers other than gcc.
2469  Updated projects/visualc6 and visualc71 with "/d PNG_CONFIGURE_LIBPNG".
2470    They should work but still need to be updated to remove
2471    references to pnggccrd.c or pngvcrd.c and ASM building.
2472  Added README.txt to the beos, cbuilder5, netware, and xcode projects warning
2473    that they need to be updated, to remove references to pnggccrd.c and
2474    pngvcrd.c and to depend on pngpriv.h
2475  Removed three direct references to read_info_ptr members in pngtest.c
2476    that were detected by the new PNG_DEPSTRUCT macro.
2477  Moved the png_debug macro definitions and the png_read_destroy(),
2478    png_write_destroy() and png_far_to_near() prototypes from png.h
2479    to pngpriv.h (John Bowler)
2480  Moved the synopsis lines for png_read_destroy(), png_write_destroy()
2481    png_debug(), png_debug1(), and png_debug2() from libpng.3 to libpngpf.3.
2482
2483Version 1.4.0beta94 [November 9, 2009]
2484  Removed the obsolete, unused pnggccrd.c and pngvcrd.c files.
2485  Updated CMakeLists.txt to add "-DPNG_CONFIGURE_LIBPNG" to the definitions.
2486  Removed dependency of pngtest.o on pngpriv.h in the makefiles.
2487  Only #define PNG_DEPSTRUCT, etc. in pngconf.h if not already defined.
2488
2489Version 1.4.0beta95 [November 10, 2009]
2490  Changed png_check_sig() to !png_sig_cmp() in contrib programs.
2491  Added -DPNG_CONFIGURE_LIBPNG to contrib/pngminm/*/makefile
2492  Changed png_check_sig() to !png_sig_cmp() in contrib programs.
2493  Corrected the png_get_IHDR() call in contrib/gregbook/readpng2.c
2494  Changed pngminim/*/gather.sh to stop trying to remove pnggccrd.c and pngvcrd.c
2495  Added dependency on pngpriv.h in contrib/pngminim/*/makefile
2496
2497Version 1.4.0beta96 [November 12, 2009]
2498  Renamed scripts/makefile.wce to scripts/makefile.cegcc
2499  Revised Makefile.am to use libpng.sys while building libpng.so
2500    so that only PNG_EXPORT functions are exported.
2501  Removed the deprecated png_check_sig() function/macro.
2502  Removed recently removed function names from scripts/*.def
2503  Revised pngtest.png to put chunks in the same order written by pngtest
2504    (evidently the same change made in libpng-1.0beta54 was lost).
2505  Added PNG_PRIVATE macro definition in pngconf.h for possible future use.
2506
2507Version 1.4.0beta97 [November 13, 2009]
2508  Restored pngtest.png to the libpng-1.4.0beta7 version.
2509  Removed projects/beos and netware.txt; no one seems to be supporting them.
2510  Revised Makefile.in
2511
2512Version 1.4.0beta98 [November 13, 2009]
2513  Added the "xcode" project to zip distributions,
2514  Fixed a typo in scripts/pngwin.def introduced in beta97.
2515
2516Version 1.4.0beta99 [November 14, 2009]
2517  Moved libpng-config.in and libpng.pc-configure.in out of the scripts
2518    directory, to libpng-config.in and libpng-pc.in, respectively, and
2519    modified Makefile.am and configure.ac accordingly.  Now "configure"
2520    needs nothing from the "scripts" directory.
2521  Avoid redefining PNG_CONST in pngconf.h
2522
2523Version 1.4.0beta100 [November 14, 2009]
2524  Removed ASM builds from projects/visualc6 and projects/visualc71
2525  Removed scripts/makefile.nommx and makefile.vcawin32
2526  Revised CMakeLists.txt to account for new location of libpng-config.in
2527    and libpng-pc.in
2528  Updated INSTALL to reflect removal and relocation of files.
2529
2530Version 1.4.0beta101 [November 14, 2009]
2531  Restored the binary files (*.jpg, *.png, some project files) that were
2532    accidentally deleted from the zip and 7z distributions when the xcode
2533    project was added.
2534
2535Version 1.4.0beta102 [November 18, 2009]
2536  Added libpng-config.in and libpng-pc.in to the zip and 7z distributions.
2537  Fixed a typo in projects/visualc6/pngtest.dsp, introduced in beta100.
2538  Moved descriptions of makefiles and other scripts out of INSTALL into
2539    scripts/README.txt
2540  Updated the copyright year in scripts/pngwin.rc from 2006 to 2009.
2541
2542Version 1.4.0beta103 [November 21, 2009]
2543  Removed obsolete comments about ASM from projects/visualc71/README_zlib.txt
2544  Align row_buf on 16-byte boundary in memory.
2545  Restored the PNG_WRITE_FLUSH_AFTER_IEND_SUPPORTED guard around the call
2546    to png_flush() after png_write_IEND().  See 1.4.0beta32, 1.4.0beta50
2547    changes above and 1.2.30, 1.2.30rc01 and rc03 in 1.2.41 CHANGES.  Someone
2548    needs this feature.
2549  Make the 'png_jmpbuf' macro expand to a call that records the correct
2550    longjmp function as well as returning a pointer to the setjmp
2551    jmp_buf buffer, and marked direct access to jmpbuf 'deprecated'.
2552    (John Bowler)
2553
2554Version 1.4.0beta104 [November 22, 2009]
2555  Removed png_longjmp_ptr from scripts/*.def and libpng.3
2556  Rebuilt configure scripts with autoconf-2.65
2557
2558Version 1.4.0beta105 [November 25, 2009]
2559  Use fast integer PNG_DIVIDE_BY_255() or PNG_DIVIDE_BY_65535()
2560    to accomplish alpha premultiplication when
2561    PNG_READ_COMPOSITE_NODIV_SUPPORTED is defined.
2562  Changed "/255" to "/255.0" in background calculations to make it clear
2563    that the 255 is used as a double.
2564
2565Version 1.4.0beta106 [November 27, 2009]
2566  Removed premultiplied alpha feature.
2567
2568Version 1.4.0beta107 [December 4, 2009]
2569  Updated README
2570  Added "#define PNG_NO_PEDANTIC_WARNINGS" in the libpng source files.
2571  Removed "-DPNG_CONFIGURE_LIBPNG" from the makefiles and projects.
2572  Revised scripts/makefile.netbsd, makefile.openbsd, and makefile.sco
2573    to put png.h and pngconf.h in $prefix/include, like the other scripts,
2574    instead of in $prefix/include/libpng.  Also revised makefile.sco
2575    to put them in $prefix/include/libpng15 instead of in
2576    $prefix/include/libpng/libpng15.
2577
2578Version 1.4.0beta108 [December 11, 2009]
2579  Removed leftover "-DPNG_CONFIGURE_LIBPNG" from contrib/pngminim/*/makefile
2580  Relocated png_do_chop() to its original position in pngrtran.c; the
2581    change in version 1.2.41beta08 caused transparency to be handled wrong
2582    in some 16-bit datastreams (Yusaku Sugai).
2583
2584Version 1.4.0beta109 [December 13, 2009]
2585  Added "bit_depth" parameter to the private png_build_gamma_table() function.
2586  Pass bit_depth=8 to png_build_gamma_table() when bit_depth is 16 but the
2587    PNG_16_TO_8 transform has been set, to avoid unnecessary build of 16-bit
2588    tables.
2589
2590Version 1.4.0rc02 [December 20, 2009]
2591  Declared png_cleanup_needed "volatile" in pngread.c and pngwrite.c
2592
2593Version 1.4.0rc03 [December 22, 2009]
2594  Renamed libpng-pc.in back to libpng.pc.in and revised CMakeLists.txt
2595    (revising the change in 1.4.0beta99)
2596
2597Version 1.4.0rc04 [December 25, 2009]
2598  Swapped PNG_UNKNOWN_CHUNKS_SUPPORTED and PNG_HANDLE_AS_UNKNOWN_SUPPORTED
2599    in pngset.c to be consistent with other changes in version 1.2.38.
2600
2601Version 1.4.0rc05 [December 25, 2009]
2602  Changed "libpng-pc.in" to "libpng.pc.in" in configure.ac, configure, and
2603    Makefile.in to be consistent with changes in libpng-1.4.0rc03
2604
2605Version 1.4.0rc06 [December 29, 2009]
2606  Reverted the gamma_table changes from libpng-1.4.0beta109.
2607  Fixed some indentation errors.
2608
2609Version 1.4.0rc07 [January 1, 2010]
2610  Revised libpng*.txt and libpng.3 about 1.2.x->1.4.x differences.
2611  Use png_calloc() instead of png_malloc(); png_memset() in pngrutil.c
2612  Update copyright year to 2010.
2613
2614Version 1.4.0rc08 [January 2, 2010]
2615  Avoid deprecated references to png_ptr-io_ptr and png_ptr->error_ptr
2616    in pngtest.c
2617
2618Version 1.4.0 [January 3, 2010]
2619  No changes.
2620
2621Version 1.4.1beta01 [January 8, 2010]
2622  Updated CMakeLists.txt for consistent indentation and to avoid an
2623    unclosed if-statement warning (Philip Lowman).
2624  Revised Makefile.am and Makefile.in to remove references to Y2KINFO,
2625    KNOWNBUG, and libpng.la (Robert Schwebel).
2626  Revised the makefiles to install the same files and symbolic
2627    links as configure, except for libpng.la and libpng14.la.
2628  Make png_set|get_compression_buffer_size() available even when
2629    PNG_WRITE_SUPPORTED is not enabled.
2630  Revised Makefile.am and Makefile.in to simplify their maintenance.
2631  Revised scripts/makefile.linux to install a link to libpng14.so.14.1
2632
2633Version 1.4.1beta02 [January 9, 2010]
2634  Revised the rest of the makefiles to install a link to libpng14.so.14.1
2635
2636Version 1.4.1beta03 [January 10, 2010]
2637  Removed png_set_premultiply_alpha() from scripts/*.def
2638
2639Version 1.4.1rc01 [January 16, 2010]
2640  No changes.
2641
2642Version 1.4.1beta04 [January 23, 2010]
2643  Revised png_decompress_chunk() to improve speed and memory usage when
2644    decoding large chunks.
2645  Added png_set|get_chunk_malloc_max() functions.
2646
2647Version 1.4.1beta05 [January 26, 2010]
2648  Relocated "int k" declaration in pngtest.c to minimize its scope.
2649
2650Version 1.4.1beta06 [January 28, 2010]
2651  Revised png_decompress_chunk() to use a two-pass method suggested by
2652    John Bowler.
2653
2654Version 1.4.1beta07 [February 6, 2010]
2655  Folded some long lines in the source files.
2656  Added defineable PNG_USER_CHUNK_CACHE_MAX, PNG_USER_CHUNK_MALLOC_MAX,
2657    and a PNG_USER_LIMITS_SUPPORTED flag.
2658  Eliminated use of png_ptr->irowbytes and reused the slot in png_ptr as
2659    png_ptr->png_user_chunk_malloc_max.
2660  Revised png_push_save_buffer() to do fewer but larger png_malloc() calls.
2661
2662Version 1.4.1beta08 [February 6, 2010]
2663  Minor cleanup and updating of dates and copyright year.
2664
2665Version 1.5.0beta01 [February 7, 2010]
2666  Moved declaration of png_struct into private pngstruct.h and png_info
2667    into pnginfo.h
2668
2669Version 1.4.1beta09 and 1.5.0beta02 [February 7, 2010]
2670  Reverted to original png_push_save_buffer() code.
2671
2672Version 1.4.1beta10 and 1.5.0beta03 [February 8, 2010]
2673  Return allocated "old_buffer" in png_push_save_buffer() before
2674    calling png_error(), to avoid a potential memory leak.
2675  Updated configure script to use SO number 15.
2676
2677Version 1.5.0beta04 [February 9, 2010]
2678  Removed malformed "incomplete struct declaration" of png_info from png.h
2679
2680Version 1.5.0beta05 [February 12, 2010]
2681  Removed PNG_DEPSTRUCT markup in pngstruct.h and pnginfo.h, and undid the
2682    linewrapping that it entailed.
2683  Revised comments in pngstruct.h and pnginfo.h and added pointers to
2684    the libpng license.
2685  Changed PNG_INTERNAL to PNG_EXPOSE_INTERNAL_STRUCTURES
2686  Removed the cbuilder5 project, which has not been updated to 1.4.0.
2687
2688Version 1.4.1beta12 and 1.5.0beta06 [February 14, 2010]
2689  Fixed type declaration of png_get_chunk_malloc_max() in pngget.c (Daisuke
2690    Nishikawa)
2691
2692Version 1.5.0beta07 [omitted]
2693
2694Version 1.5.0beta08 [February 19, 2010]
2695  Changed #ifdef PNG_NO_STDIO_SUPPORTED to #ifdef PNG_NO_CONSOLE_IO_SUPPORTED
2696    wherever png_snprintf() is used to construct error and warning messages.
2697  Noted in scripts/makefile.mingw that it expects to be run under MSYS.
2698  Removed obsolete unused MMX-querying support from contrib/gregbook
2699  Added exported png_longjmp() function.
2700  Removed the AIX redefinition of jmpbuf in png.h
2701  Added -D_ALLSOURCE in configure.ac, makefile.aix, and CMakeLists.txt
2702    when building on AIX.
2703
2704Version 1.5.0beta09 [February 19, 2010]
2705  Removed -D_ALLSOURCE from configure.ac, makefile.aix, and CMakeLists.txt.
2706  Changed the name of png_ptr->jmpbuf to png_ptr->png_jmpbuf in pngstruct.h
2707
2708Version 1.5.0beta10 [February 25, 2010]
2709  Removed unused gzio.c from contrib/pngminim gather and makefile scripts
2710  Removed replacement error handlers from contrib/gregbook.  Because of
2711    the new png_longjmp() function they are no longer needed.
2712
2713Version 1.5.0beta11 [March 6, 2010]
2714  Removed checking for already-included setjmp.h from pngconf.h
2715  Fixed inconsistent indentations and made numerous cosmetic changes.
2716  Revised the "SEE ALSO" style of libpng.3, libpngpf.3, and png.5
2717
2718Version 1.5.0beta12 [March 9, 2010]
2719  Moved "#include png.h" inside pngpriv.h and removed "#include png.h" from
2720    the source files, along with "#define PNG_EXPOSE_INTERNAL_STRUCTURES"
2721    and "#define PNG_NO_PEDANTIC_WARNINGS" (John Bowler).
2722  Created new pngdebug.h and moved debug definitions there.
2723
2724Version 1.5.0beta13 [March 10, 2010]
2725  Protect pngstruct.h, pnginfo.h, and pngdebug.h from being included twice.
2726  Revise the "#ifdef" blocks in png_inflate() so it will compile when neither
2727    PNG_USER_CHUNK_MALLOC_MAX nor PNG_SET_CHUNK_MALLOC_LIMIT_SUPPORTED
2728    is defined.
2729  Removed unused png_measure_compressed_chunk() from pngpriv.h and libpngpf.3
2730  Moved the 'config.h' support from pngconf.h to pngpriv.h
2731  Removed PNGAPI from the png_longjmp_ptr typedef.
2732  Eliminated dependence of pngtest.c on the private pngdebug.h file.
2733  Make all png_debug macros into *unterminated* statements or
2734    expressions (i.e. a trailing ';' must always be added) and correct
2735    the format statements in various png_debug messages.
2736
2737Version 1.5.0beta14 [March 14, 2010]
2738  Removed direct access to png_ptr->io_ptr from the Windows code in pngtest.c
2739  Revised Makefile.am to account for recent additions and replacements.
2740  Corrected CE and OS/2 DEF files (scripts/png*def) for symbols removed and
2741    added ordinal numbers to the Windows DEF file and corrected the duplicated
2742    ordinal numbers on CE symbols that are commented out.
2743  Added back in export symbols that can be present in the Windows build but
2744    are disabled by default.
2745  PNG_EXPORT changed to include an 'ordinal' field for DEF file generation.
2746    PNG_CALLBACK added to make callback definitions uniform.  PNGAPI split
2747    into PNGCAPI (base C form), PNGAPI (exports) and PNGCBAPI (callbacks),
2748    and appropriate changes made to all files.  Cygwin builds re-hinged to
2749    allow procedure call standard changes and to remove the need for the DEF
2750    file (fixes build on Cygwin).
2751  Enabled 'attribute' warnings that are relevant to library APIs and callbacks.
2752  Changed rules for generation of the various symbol files and added a new
2753    rule for a DEF file (which is also added to the distribution).
2754  Updated the symbol file generation to stop it adding spurious spaces
2755    to EOL (coming from preprocessor macro expansion).  Added a facility
2756    to join tokens in the output and rewrite *.dfn to use this.
2757  Eliminated scripts/*.def in favor of libpng.def; updated projects/visualc71
2758    and removed scripts/makefile.cygwin.
2759  Made PNG_BUILD_DLL safe: it can be set whenever a DLL is being built.
2760  Removed the include of sys/types.h - apparently unnecessary now on the
2761    platforms on which it happened (all but Mac OS and RISC OS).
2762  Moved the Mac OS test into pngpriv.h (the only place it is used.)
2763
2764Version 1.5.0beta15 [March 17, 2010]
2765  Added symbols.chk target to Makefile.am to validate the symbols in png.h
2766    against the new DEF file scripts/symbols.def.
2767  Changed the default DEF file back to pngwin.def.
2768  Removed makefile.mingw.
2769  Eliminated PNG_NO_EXTERN and PNG_ALL_EXTERN
2770
2771Version 1.5.0beta16 [April 1, 2010]
2772  Make png_text_struct independent of PNG_iTXt_SUPPORTED, so that
2773    fields are initialized in all configurations.  The READ/WRITE
2774    macros (PNG_(READ|WRITE)_iTXt_SUPPORTED) still function as
2775    before to disable code to actually read or write iTXt chunks
2776    and iTXt_SUPPORTED can be used to detect presence of either
2777    read or write support (but it is probably better to check for
2778    the one actually required - read or write.)
2779  Combined multiple png_warning() calls for a single error.
2780  Restored the macro definition of png_check_sig().
2781
2782Version 1.5.0beta17 [April 17, 2010]
2783  Added some "(long)" typecasts to printf calls in png_handle_cHRM().
2784  Documented the fact that png_set_dither() was disabled since libpng-1.4.0.
2785  Reenabled png_set_dither() but renamed it to png_set_quantize() to reflect
2786    more accurately what it actually does.  At the same time, renamed
2787    the PNG_DITHER_[RED,GREEN_BLUE]_BITS macros to
2788    PNG_QUANTIZE_[RED,GREEN,BLUE]_BITS.
2789  Added some "(long)" typecasts to printf calls in png_handle_cHRM().
2790  Freeze build-time only configuration in the build.
2791    In all prior versions of libpng most configuration options
2792    controlled by compiler #defines had to be repeated by the
2793    application code that used libpng.  This patch changes this
2794    so that compilation options that can only be changed at build
2795    time are frozen in the build.  Options that are compiler
2796    dependent (and those that are system dependent) are evaluated
2797    each time - pngconf.h holds these.  Options that can be changed
2798    per-file in the application are in png.h.  Frozen options are
2799    in the new installed header file pnglibconf.h (John Bowler)
2800  Removed the xcode project because it has not been updated to work
2801    with libpng-1.5.0.
2802  Removed the ability to include optional pngusr.h
2803
2804Version 1.5.0beta18 [April 17, 2010]
2805  Restored the ability to include optional pngusr.h
2806  Moved replacements for png_error() and png_warning() from the
2807    contrib/pngminim project to pngerror.c, for use when warnings or
2808    errors are disabled via PNG_NO_WARN or PNG_NO_ERROR_TEXT, to avoid
2809    storing unneeded error/warning text.
2810  Updated contrib/pngminim project to work with the new pnglibconf.h
2811  Added some PNG_NO_* defines to contrib/pngminim/*/pngusr.h to save space.
2812
2813Version 1.5.0beta19 [April 24, 2010]
2814  Added PNG_{READ,WRITE}_INT_FUNCTIONS_SUPPORTED.  This allows the functions
2815    to read and write ints to be disabled independently of PNG_USE_READ_MACROS,
2816    which allows libpng to be built with the functions even though the default
2817    is to use the macros - this allows applications to choose at app build
2818    time whether or not to use macros (previously impossible because the
2819    functions weren't in the default build.)
2820  Changed Windows calling convention back to __cdecl for API functions.
2821    For Windows/x86 platforms only:
2822      __stdcall is no longer needed for Visual Basic, so libpng-1.5.0 uses
2823      __cdecl throughout (both API functions and callbacks) on Windows/x86
2824      platforms.
2825  Replaced visualc6 and visualc71 projects with new vstudio project
2826  Relaxed the overly-restrictive permissions of some files.
2827
2828Version 1.5.0beta20 [April 24, 2010]
2829  Relaxed more overly-restrictive permissions of some files.
2830
2831Version 1.5.0beta21 [April 27, 2010]
2832  Removed some unwanted binary bytes and changed CRLF to NEWLINE in the new
2833    vstudio project files, and some trivial editing of some files in the
2834    scripts directory.
2835  Set PNG_NO_READ_BGR, PNG_NO_IO_STATE, and PNG_NO_TIME_RFC1123 in
2836    contrib/pngminim/decoder/pngusr.h to make a smaller decoder application.
2837
2838Version 1.5.0beta22 [April 28, 2010]
2839  Fixed dependencies of GET_INT_32 - it does not require READ_INT_FUNCTIONS
2840    because it has a macro equivalent.
2841  Improved the options.awk script; added an "everything off" option.
2842  Revised contrib/pngminim to use the "everything off" option in pngusr.dfa.
2843
2844Version 1.5.0beta23 [April 29, 2010]
2845  Corrected PNG_REMOVED macro to take five arguments.
2846    The macro was documented with two arguments (name,ordinal), however
2847    the symbol checking .dfn files assumed five arguments.  The five
2848    argument form seems more useful so it is changed to that.
2849  Corrected PNG_UNKNOWN_CHUNKS_SUPPORTED to PNG_HANDLE_AS_UNKNOWN_SUPPORTED
2850    in gregbook/readpng2.c
2851  Corrected protection of png_get_user_transform_ptr. The API declaration in
2852    png.h is removed if both READ and WRITE USER_TRANSFORM are turned off
2853    but was left defined in pngtrans.c
2854  Added logunsupported=1 to cause pnglibconf.h to document disabled options.
2855    This makes the installed pnglibconf.h more readable but causes no
2856    other change.  The intention is that users of libpng will find it
2857    easier to understand if an API they need is missing.
2858  Include png_reset_zstream() in png.c only when PNG_READ_SUPPORTED is defined.
2859  Removed dummy_inflate.c from contrib/pngminim/encoder
2860  Removed contrib/pngminim/*/gather.sh; gathering is now done in the makefile.
2861
2862Version 1.5.0beta24 [May 7, 2010]
2863  Use bitwise "&" instead of arithmetic mod in pngrutil.c calculation of the
2864    offset of the png_ptr->rowbuf pointer into png_ptr->big_row_buf.
2865  Added more blank lines for readability.
2866
2867Version 1.5.0beta25 [June 18, 2010]
2868  In pngpread.c: png_push_have_row() add check for new_row > height
2869  Removed the now-redundant check for out-of-bounds new_row from example.c
2870
2871Version 1.5.0beta26 [June 18, 2010]
2872  In pngpread.c: png_push_process_row() add check for too many rows.
2873
2874Version 1.5.0beta27 [June 18, 2010]
2875  Removed the check added in beta25 as it is now redundant.
2876
2877Version 1.5.0beta28 [June 20, 2010]
2878  Rewrote png_process_IDAT_data to consistently treat extra data as warnings
2879    and handle end conditions more cleanly.
2880  Removed the new (beta26) check in png_push_process_row().
2881
2882Version 1.5.0beta29 [June 21, 2010]
2883  Revised scripts/options.awk to work on Sunos (but still doesn't work)
2884  Added comment to options.awk and contrib/pngminim/*/makefile to try nawk.
2885
2886Version 1.5.0beta30 [June 22, 2010]
2887  Stop memory leak when reading a malformed sCAL chunk.
2888
2889Version 1.5.0beta31 [June 26, 2010]
2890  Revised pngpread.c patch of beta28 to avoid an endless loop.
2891  Removed some trailing blanks.
2892
2893Version 1.5.0beta32 [June 26, 2010]
2894  Removed leftover scripts/options.patch and scripts/options.rej
2895
2896Version 1.5.0beta33 [July 6, 3010]
2897  Made FIXED and FLOATING options consistent in the APIs they enable and
2898    disable.  Corrected scripts/options.awk to handle both command line
2899    options and options specified in the .dfa files.
2900  Changed char *msg to PNG_CONST char *msg in pngrutil.c
2901  Make png_set_sRGB_gAMA_and_cHRM set values using either the fixed or
2902    floating point APIs, but not both.
2903  Reversed patch to remove error handler when the jmp_buf is stored in the
2904    main program structure, not the png_struct.
2905    The error handler is needed because the default handler in libpng will
2906    always use the jmp_buf in the library control structure; this is never
2907    set.  The gregbook code is a useful example because, even though it
2908    uses setjmp/longjmp, it shows how error handling can be implemented
2909    using control mechanisms not directly supported by libpng.  The
2910    technique will work correctly with mechanisms such as Microsoft
2911    Structure Exceptions or C++ exceptions (compiler willing - note that gcc
2912    does not by default support interworking of C and C++ error handling.)
2913  Reverted changes to call png_longjmp in contrib/gregbook where it is not
2914    appropriate.  If mainprog->jmpbuf is used by setjmp, then png_longjmp
2915    cannot be used.
2916  Changed "extern PNG_EXPORT" to "PNG_EXPORT" in png.h (Jan Nijtmans)
2917  Changed "extern" to "PNG_EXTERN" in pngpriv.h (except for the 'extern "C" {')
2918
2919Version 1.5.0beta34 [July 12, 2010]
2920  Put #ifndef PNG_EXTERN, #endif around the define PNG_EXTERN in pngpriv.h
2921
2922Version 1.5.0beta35 [July 24, 2010]
2923  Removed some newly-added TAB characters.
2924  Added -DNO_PNG_SNPRINTF to CFLAGS in scripts/makefile.dj2
2925  Moved the definition of png_snprintf() outside of the enclosing
2926    #ifdef blocks in pngconf.h
2927
2928Version 1.5.0beta36 [July 29, 2010]
2929  Patches by John Bowler:
2930  Fixed point APIs are now supported throughout (no missing APIs).
2931  Internal fixed point arithmetic support exists for all internal floating
2932    point operations.
2933  sCAL validates the floating point strings it is passed.
2934  Safe, albeit rudimentary, Watcom support is provided by PNG_API_RULE==2
2935  Two new APIs exist to get the number of passes without turning on the
2936    PNG_INTERLACE transform and to get the number of rows in the current
2937    pass.
2938  A new test program, pngvalid.c, validates the gamma code.
2939  Errors in the 16-bit gamma correction (overflows) have been corrected.
2940  cHRM chunk testing is done consistently (previously the floating point
2941    API bypassed it, because the test really didn't work on FP, now the test
2942    is performed on the actual values to be stored in the PNG file so it
2943    works in the FP case too.)
2944  Most floating point APIs now simply call the fixed point APIs after
2945    converting the values to the fixed point form used in the PNG file.
2946  The standard headers no longer include zlib.h, which is currently only
2947    required for pngstruct.h and can therefore be internal.
2948  Revised png_get_int_32 to undo the PNG two's complement representation of
2949    negative numbers.
2950
2951Version 1.5.0beta37 [July 30, 2010]
2952  Added a typecast in png_get_int_32() in png.h and pngrutil.h to avoid
2953    a compiler warning.
2954  Replaced oFFs 0,0 with oFFs -10,20 in pngtest.png
2955
2956Version 1.5.0beta38 [July 31, 2010]
2957  Implemented remaining "_fixed" functions.
2958  Corrected a number of recently introduced warnings mostly resulting from
2959    safe but uncast assignments to shorter integers.  Also added a zlib
2960    VStudio release library project because the latest zlib Official Windows
2961    build does not include such a thing.
2962  Revised png_get_int_16() to be similar to png_get_int_32().
2963  Restored projects/visualc71.
2964
2965Version 1.5.0beta39 [August 2, 2010]
2966  VisualC/GCC warning fixes, VisualC build fixes
2967  The changes include support for function attributes in VC in addition to
2968    those already present in GCC - necessary because without these some
2969    warnings are unavoidable.  Fixes include signed/unsigned fixes in
2970    pngvalid and checks with gcc -Wall -Wextra -Wunused.
2971  VC requires function attributes on function definitions as well as
2972    declarations, PNG_FUNCTION has been added to enable this and the
2973    relevant function definitions changed.
2974
2975Version 1.5.0beta40 [August 6, 2010]
2976  Correct use of _WINDOWS_ in pngconf.h
2977  Removed png_mem_ #defines; they are no longer used.
2978  Added the sRGB chunk to pngtest.png
2979
2980Version 1.5.0beta41 [August 11, 2010]
2981  Added the cHRM chunk to pngtest.png
2982  Don't try to use version-script with cygwin/mingw.
2983  Revised contrib/gregbook to work under cygwin/mingw.
2984
2985Version 1.5.0beta42 [August 18, 2010]
2986  Add .dll.a to the list of extensions to be symlinked by Makefile.am (Yaakov)
2987  Made all API functions that have const arguments and constant string
2988    literal pointers declare them (John Bowler).
2989
2990Version 1.5.0beta43 [August 20, 2010]
2991  Removed spurious tabs, shorten long lines (no source change)
2992    Also added scripts/chkfmt to validate the format of all the files that can
2993    reasonably be validated (it is suggested to run "make distclean" before
2994    checking, because some machine generated files have long lines.)
2995  Reformatted the CHANGES file to be more consistent throughout.
2996  Made changes to address various issues identified by GCC, mostly
2997    signed/unsigned and shortening problems on assignment but also a few
2998    difficult to optimize (for GCC) loops.
2999  Fixed non-GCC fixed point builds.  In png.c a declaration was misplaced
3000    in an earlier update.  Fixed to declare the auto variables at the head.
3001  Use cexcept.h in pngvalid.c.
3002
3003Version 1.5.0beta44 [August 24, 2010]
3004  Updated CMakeLists.txt to use CMAKE_INSTALL_LIBDIR variable; useful for
3005    installing libpng in /usr/lib64 (Funda Wang).
3006  Revised CMakeLists.txt to put the man pages in share/man/man* not man/man*
3007  Revised CMakeLists.txt to make symlinks instead of copies when installing.
3008  Changed PNG_LIB_NAME from pngNN to libpngNN in CMakeLists.txt (Philip Lowman)
3009  Implemented memory checks within pngvalid
3010  Reformatted/rearranged pngvalid.c to assist use of progressive reader.
3011  Check interlaced images in pngvalid
3012  Clarified pngusr.h comments in pnglibconf.dfa
3013  Simplified the pngvalid error-handling code now that cexcept.h is in place.
3014  Implemented progressive reader in pngvalid.c for standard tests
3015  Implemented progressive read in pngvalid.c gamma tests
3016  Turn on progressive reader in pngvalid.c by default and tidy code.
3017
3018Version 1.5.0beta45 [August 26, 2010]
3019  Added an explicit make step to projects/vstudio for pnglibconf.h
3020    Also corrected zlib.vcxproj into which Visual Studio had introduced
3021    what it calls an "authoring error".  The change to make pnglibconf.h
3022    simply copies the file; in the future it may actually generate the
3023    file from scripts/pnglibconf.dfa as the other build systems do.
3024  Changed pngvalid to work when floating point APIs are disabled
3025  Renamed the prebuilt scripts/pnglibconf.h to scripts/pnglibconf.h.prebuilt
3026  Supply default values for PNG_USER_PRIVATEBUILD and PNG_USER_DLLFNAME_POSTFIX
3027    in pngpriv.h in case the user neglected to define them in their pngusr.h
3028
3029Version 1.5.0beta46 [August 28, 2010]
3030  Added new private header files to libpng_sources in CMakeLists.txt
3031  Added PNG_READ_16BIT, PNG_WRITE_16BIT, and PNG_16BIT options.
3032  Added reference to scripts/pnglibconf.h.prebuilt in the visualc71 project.
3033
3034Version 1.5.0beta47 [September 11, 2010]
3035  Fixed a number of problems with 64-bit compilation reported by Visual
3036    Studio 2010 (John Bowler).
3037
3038Version 1.5.0beta48 [October 4, 2010]
3039  Updated CMakeLists.txt (Philip Lowman).
3040  Revised autogen.sh to recognize and use $AUTOCONF, $AUTOMAKE, $AUTOHEADER,
3041    $AUTOPOINT, $ACLOCAL and $LIBTOOLIZE
3042  Fixed problem with symbols creation in Makefile.am which was assuming that
3043    all versions of ccp write to standard output by default (Martin Banky). The
3044    bug was introduced in libpng-1.2.9beta5.
3045  Removed unused mkinstalldirs.
3046
3047Version 1.5.0beta49 [October 8, 2010]
3048  Undid Makefile.am revision of 1.5.0beta48.
3049
3050Version 1.5.0beta50 [October 14, 2010]
3051  Revised Makefile.in to account for mkinstalldirs being removed.
3052  Added some "(unsigned long)" typecasts in printf statements in pngvalid.c.
3053  Suppressed a compiler warning in png_handle_sPLT().
3054  Check for out-of-range text compression mode in png_set_text().
3055
3056Version 1.5.0beta51 [October 15, 2010]
3057  Changed embedded dates to "(PENDING RELEASE) in beta releases (and future
3058    rc releases) to minimize the difference between releases.
3059
3060Version 1.5.0beta52 [October 16, 2010]
3061  Restored some of the embedded dates (in png.h, png.c, documentation, etc.)
3062
3063Version 1.5.0beta53 [October 18, 2010]
3064  Updated INSTALL to mention using "make maintainer-clean" and to remove
3065    obsolete statement about a custom ltmain.sh
3066  Disabled "color-tests" by default in Makefile.am so it will work with
3067    automake versions earlier than 1.11.1
3068  Use document name "libpng-manual.txt" instead of "libpng-<version>.txt"
3069    to simplify version differences.
3070  Removed obsolete remarks about setjmp handling from INSTALL.
3071  Revised and renamed the typedef in png.h and png.c that was designed
3072    to catch library and header mismatch.
3073
3074Version 1.5.0beta54 [November 10, 2010]
3075  Require 48 bytes, not 64 bytes, for big_row_buf in overflow checks.
3076  Used a consistent structure for the pngget.c functions.
3077
3078Version 1.5.0beta55 [November 21, 2010]
3079  Revised png_get_uint_32, png_get_int_32, png_get_uint_16 (Cosmin)
3080  Moved reading of file signature into png_read_sig (Cosmin)
3081  Fixed atomicity of chunk header serialization (Cosmin)
3082  Added test for io_state in pngtest.c (Cosmin)
3083  Added "#!/bin/sh" at the top of contrib/pngminim/*/gather.sh scripts.
3084  Changes to remove gcc warnings (John Bowler)
3085    Certain optional gcc warning flags resulted in warnings in libpng code.
3086    With these changes only -Wconversion and -Wcast-qual cannot be turned on.
3087    Changes are trivial rearrangements of code.  -Wconversion is not possible
3088    for pngrutil.c (because of the widespread use of += et al on variables
3089    smaller than (int) or (unsigned int)) and -Wcast-qual is not possible
3090    with pngwio.c and pngwutil.c because the 'write' callback and zlib
3091    compression both fail to declare their input buffers with 'const'.
3092
3093Version 1.5.0beta56 [December 7, 2010]
3094  Added the private PNG_UNUSED() macro definition in pngpriv.h.
3095  Added some commentary about PNG_EXPORT in png.h and pngconf.h
3096  Revised PNG_EXPORT() macro and added PNG_EXPORTA() macro, with the
3097    objective of simplifying and improving the cosmetic appearance of png.h.
3098  Fixed some incorrect "=" macro names in pnglibconf.dfa
3099  Included documentation of changes in 1.5.0 from 1.4.x in libpng-manual.txt
3100
3101Version 1.5.0beta57 [December 9, 2010]
3102  Documented the pngvalid gamma error summary with additional comments and
3103    print statements.
3104  Improved missing symbol handling in checksym.awk; symbols missing in both
3105    the old and new files can now be optionally ignored, treated as errors
3106    or warnings.
3107  Removed references to pngvcrd.c and pnggccrd.c from the vstudio project.
3108  Updated "libpng14" to "libpng15" in the visualc71 project.
3109  Enabled the strip16 tests in pngvalid.`
3110  Don't display test results (except PASS/FAIL) when running "make test".
3111    Instead put them in pngtest-log.txt
3112  Added "--with-zprefix=<string>" to configure.ac
3113  Updated the prebuilt configuration files to autoconf version 2.68
3114
3115Version 1.5.0beta58 [December 19, 2010]
3116  Fixed interlace image handling and add test cases (John Bowler)
3117  Fixed the clean rule in Makefile.am to remove pngtest-log.txt
3118  Made minor changes to work around warnings in gcc 3.4
3119
3120Version 1.5.0rc01 [December 27, 2010]
3121  No changes.
3122
3123Version 1.5.0rc02 [December 27, 2010]
3124  Eliminated references to the scripts/*.def files in project/visualc71.
3125
3126Version 1.5.0rc03 [December 28, 2010]
3127  Eliminated scripts/*.def and revised Makefile.am accordingly
3128
3129Version 1.5.0rc04 [December 29, 2010]
3130  Fixed bug in background transformation handling in pngrtran.c (it was
3131    looking for the flag in png_ptr->transformations instead of in
3132    png_ptr->flags) (David Raymond).
3133
3134Version 1.5.0rc05 [December 31, 2010]
3135  Fixed typo in a comment in CMakeLists.txt (libpng14 => libpng15) (Cosmin)
3136
3137Version 1.5.0rc06 [January 4, 2011]
3138  Changed the new configure option "zprefix=string" to "zlib-prefix=string"
3139
3140Version 1.5.0rc07 [January 4, 2011]
3141  Updated copyright year.
3142
3143Version 1.5.0 [January 6, 2011]
3144  No changes.
3145
3146version 1.5.1beta01 [January 8, 2011]
3147  Added description of png_set_crc_action() to the manual.
3148  Added a note in the manual that the type of the iCCP profile was changed
3149    from png_charpp to png_bytepp in png_get_iCCP().  This change happened
3150    in version 1.5.0beta36 but is not noted in the CHANGES.  Similarly,
3151    it was changed from png_charpp to png_const_bytepp in png_set_iCCP().
3152  Ensure that png_rgb_to_gray ignores palette mapped images, if libpng
3153    internally happens to call it with one, and fixed a failure to handle
3154    palette mapped images correctly.  This fixes CVE-2690.
3155
3156Version 1.5.1beta02 [January 14, 2011]
3157  Fixed a bug in handling of interlaced images (bero at arklinux.org).
3158  Updated CMakeLists.txt (Clifford Yapp)
3159
3160Version 1.5.1beta03 [January 14, 2011]
3161  Fixed typecasting of some png_debug() statements (Cosmin)
3162
3163Version 1.5.1beta04 [January 16, 2011]
3164  Updated documentation of png_set|get_tRNS() (Thomas Klausner).
3165  Mentioned in the documentation that applications must #include "zlib.h"
3166    if they need access to anything in zlib.h, and that a number of
3167    macros such as png_memset() are no longer accessible by applications.
3168  Corrected pngvalid gamma test "sample" function to access all of the color
3169    samples of each pixel, instead of sampling the red channel three times.
3170  Prefixed variable names index, div, exp, gamma with "png_" to avoid "shadow"
3171    warnings, and (mistakenly) changed png_exp() to exp().
3172
3173Version 1.5.1beta05 [January 16, 2011]
3174  Changed variable names png_index, png_div, png_exp, and png_gamma to
3175    char_index, divisor, exp_b10, and gamma_val, respectively, and
3176    changed exp() back to png_exp().
3177
3178Version 1.5.1beta06 [January 20, 2011]
3179  Prevent png_push_crc_skip() from hanging while reading an unknown chunk
3180    or an over-large compressed zTXt chunk with the progressive reader.
3181  Eliminated more GCC "shadow" warnings.
3182  Revised png_fixed() in png.c to avoid compiler warning about reaching the
3183    end without returning anything.
3184
3185Version 1.5.1beta07 [January 22, 2011]
3186  In the manual, describe the png_get_IHDR() arguments in the correct order.
3187  Added const_png_structp and const_png_infop types, and used them in
3188    prototypes for most png_get_*() functions.
3189
3190Version 1.5.1beta08 [January 23, 2011]
3191  Added png_get_io_chunk_type() and deprecated png_get_io_chunk_name()
3192  Added synopses for the IO_STATE functions and other missing synopses
3193    to the manual. Removed the synopses from libpngpf.3 because they
3194    were out of date and no longer useful.  Better information can be
3195    obtained by reading the prototypes and comments in pngpriv.h
3196  Attempted to fix cpp on Solaris with S. Studio 12 cc, fix build
3197    Added a make macro DFNCPP that is a CPP that will accept the tokens in
3198    a .dfn file and adds configure stuff to test for such a CPP.  ./configure
3199    should fail if one is not available.
3200  Corrected const_png_ in png.h to png_const_ to avoid polluting the namespace.
3201  Added png_get_current_row_number and png_get_current_pass_number for the
3202    benefit of the user transform callback.
3203  Added png_process_data_pause and png_process_data_skip for the benefit of
3204    progressive readers that need to stop data processing or want to optimize
3205    skipping of unread data (e.g., if the reader marks a chunk to be skipped.)
3206
3207Version 1.5.1beta09 [January 24, 2011]
3208  Enhanced pngvalid, corrected an error in gray_to_rgb, corrected doc error.
3209    pngvalid contains tests of transforms, which tests are currently disabled
3210    because they are incompletely tested.  gray_to_rgb was failing to expand
3211    the bit depth for smaller bit depth images; this seems to be a long
3212    standing error and resulted, apparently, in invalid output
3213    (CVE-2011-0408, CERT VU#643140).  The documentation did not accurately
3214    describe what libpng really does when converting RGB to gray.
3215
3216Version 1.5.1beta10 [January 27, 2010]
3217  Fixed incorrect examples of callback prototypes in the manual, that were
3218    introduced in libpng-1.0.0.
3219  In addition the order of the png_get_uint macros with respect to the
3220    relevant function definitions has been reversed.  This helps the
3221    preprocessing of the symbol files be more robust.  Furthermore, the
3222    symbol file preprocessing now uses -DPNG_NO_USE_READ_MACROS even when
3223    the library may actually be built with PNG_USE_READ_MACROS; this stops
3224    the read macros interfering with the symbol file format.
3225  Made the manual, synopses, and function prototypes use the function
3226    argument names file_gamma, int_file_gamma, and srgb_intent consistently.
3227
3228Version 1.5.1beta11 [January 28, 2011]
3229  Changed PNG_UNUSED from "param=param;" to "{if(param){}}".
3230  Corrected local variable type in new API png_process_data_skip()
3231    The type was self-evidently incorrect but only causes problems on 64-bit
3232    architectures.
3233  Added transform tests to pngvalid and simplified the arguments.
3234
3235Version 1.5.1rc01 [January 29, 2011]
3236  No changes.
3237
3238Version 1.5.1rc02 [January 31, 2011]
3239  Added a request in the manual that applications do not use "png_" or
3240    "PNG_" to begin any of their own symbols.
3241  Changed PNG_UNUSED to "(void)param;" and updated the commentary in pngpriv.h
3242
3243Version 1.5.1 [February 3, 2011]
3244  No changes.
3245
3246Version 1.5.2beta01 [February 13, 2011]
3247  More -Wshadow fixes for older gcc compilers.  Older gcc versions apparently
3248    check formal parameters names in function declarations (as well as
3249    definitions) to see if they match a name in the global namespace.
3250  Revised PNG_EXPORTA macro to not use an empty parameter, to accommodate the
3251    old VisualC++ preprocessor.
3252  Turned on interlace handling in png_read_png().
3253  Fixed gcc pendantic warnings.
3254  Handle longjmp in Cygwin.
3255  Fixed png_get_current_row_number() in the interlaced case.
3256  Cleaned up ALPHA flags and transformations.
3257  Implemented expansion to 16 bits.
3258
3259Version 1.5.2beta02 [February 19, 2011]
3260  Fixed mistake in the descriptions of user read_transform and write_transform
3261    function prototypes in the manual.  The row_info struct is png_row_infop.
3262  Reverted png_get_current_row_number() to previous (1.5.2beta01) behavior.
3263  Corrected png_get_current_row_number documentation
3264  Fixed the read/write row callback documentation.
3265    This documents the current behavior, where the callback is called after
3266    every row with information pertaining to the next row.
3267
3268Version 1.5.2beta03 [March 3, 2011]
3269  Fixed scripts/makefile.vcwin32
3270  Updated contrib/pngsuite/README to add the word "modify".
3271  Define PNG_ALLOCATED to blank when _MSC_VER<1300.
3272
3273Version 1.5.2rc01 [March 19, 2011]
3274  Define remaining attributes to blank when MSC_VER<1300.
3275  ifdef out mask arrays in pngread.c when interlacing is not supported.
3276
3277Version 1.5.2rc02 [March 22, 2011]
3278  Added a hint to try CPP=/bin/cpp if "cpp -E" fails in scripts/pnglibconf.mak
3279    and in contrib/pngminim/*/makefile, eg., on SunOS 5.10, and removed "strip"
3280    from the makefiles.
3281  Fixed a bug (present since libpng-1.0.7) that makes png_handle_sPLT() fail
3282    to compile when PNG_NO_POINTER_INDEXING is defined (Chubanov Kirill)
3283
3284Version 1.5.2rc03 [March 24, 2011]
3285  Don't include standard header files in png.h while building the symbol table,
3286    to avoid cpp failure on SunOS (introduced PNG_BUILDING_SYMBOL_TABLE macro).
3287
3288Version 1.5.2 [March 31, 2011]
3289  No changes.
3290
3291Version 1.5.3beta01 [April 1, 2011]
3292  Re-initialize the zlib compressor before compressing non-IDAT chunks.
3293  Added API functions (png_set_text_compression_level() and four others) to
3294    set parameters for zlib compression of non-IDAT chunks.
3295
3296Version 1.5.3beta02 [April 3, 2011]
3297  Updated scripts/symbols.def with new API functions.
3298  Only compile the new zlib re-initializing code when text or iCCP is
3299    supported, using PNG_WRITE_COMPRESSED_TEXT_SUPPORTED macro.
3300  Improved the optimization of the zlib CMF byte (see libpng-1.2.6beta03).
3301  Optimize the zlib CMF byte in non-IDAT compressed chunks
3302
3303Version 1.5.3beta03 [April 16, 2011]
3304  Fixed gcc -ansi -pedantic compile. A strict ANSI system does not have
3305    snprintf, and the "__STRICT_ANSI__" detects that condition more reliably
3306    than __STDC__ (John Bowler).
3307  Removed the PNG_PTR_NORETURN attribute because it too dangerous. It tells
3308    the compiler that a user supplied callback (the error handler) does not
3309    return, yet there is no guarantee in practice that the application code
3310    will correctly implement the error handler because the compiler only
3311    issues a warning if there is a mistake (John Bowler).
3312  Removed the no-longer-used PNG_DEPSTRUCT macro.
3313  Updated the zlib version to 1.2.5 in the VStudio project.
3314  Fixed 64-bit builds where png_uint_32 is smaller than png_size_t in
3315    pngwutil.c (John Bowler).
3316  Fixed bug with stripping the filler or alpha channel when writing, that
3317    was introduced in libpng-1.5.2beta01 (bug report by Andrew Church).
3318
3319Version 1.5.3beta04 [April 27, 2011]
3320  Updated pngtest.png with the new zlib CMF optimization.
3321  Cleaned up conditional compilation code and of background/gamma handling
3322    Internal changes only except a new option to avoid compiling the
3323    png_build_grayscale_palette API (which is not used at all internally.)
3324    The main change is to move the transform tests (READ_TRANSFORMS,
3325    WRITE_TRANSFORMS) up one level to the caller of the APIs.  This avoids
3326    calls to spurious functions if all transforms are disabled and slightly
3327    simplifies those functions.  Pngvalid modified to handle this.
3328    A minor change is to stop the strip_16 and expand_16 interfaces from
3329    disabling each other; this allows the future alpha premultiplication
3330    code to use 16-bit intermediate values while still producing 8-bit output.
3331    png_do_background and png_do_gamma have been simplified to take a single
3332    pointer to the png_struct rather than pointers to every item required
3333    from the png_struct. This makes no practical difference to the internal
3334    code.
3335  A serious bug in the pngvalid internal routine 'standard_display_init' has
3336    been fixed - this failed to initialize the red channel and accidentally
3337    initialized the alpha channel twice.
3338  Changed png_struct jmp_buf member name from png_jmpbuf to tmp_jmpbuf to
3339    avoid a possible clash with the png_jmpbuf macro on some platforms.
3340
3341Version 1.5.3beta05 [May 6, 2011]
3342  Added the "_POSIX_SOURCE" feature test macro to ensure libpng sees the
3343    correct API. _POSIX_SOURCE is defined in pngpriv.h, pngtest.c and
3344    pngvalid.c to ensure that POSIX conformant systems disable non-POSIX APIs.
3345  Removed png_snprintf and added formatted warning messages.  This change adds
3346    internal APIs to allow png_warning messages to have parameters without
3347    requiring the host OS to implement snprintf.  As a side effect the
3348    dependency of the tIME-supporting RFC1132 code on stdio is removed and
3349    PNG_NO_WARNINGS does actually work now.
3350  Pass "" instead of '\0' to png_default_error() in png_err().  This mistake
3351    was introduced in libpng-1.2.20beta01.  This fixes CVE-2011-2691.
3352  Added PNG_WRITE_OPTIMIZE_CMF_SUPPORTED macro to make the zlib "CMF" byte
3353    optimization configureable.
3354  IDAT compression failed if preceded by a compressed text chunk (bug
3355    introduced in libpng-1.5.3beta01-02).  This was because the attempt to
3356    reset the zlib stream in png_write_IDAT happened after the first IDAT
3357    chunk had been deflated - much too late.  In this change internal
3358    functions were added to claim/release the z_stream and, hopefully, make
3359    the code more robust.  Also deflateEnd checking is added - previously
3360    libpng would ignore an error at the end of the stream.
3361
3362Version 1.5.3beta06 [May 8, 2011]
3363  Removed the -D_ALL_SOURCE from definitions for AIX in CMakeLists.txt
3364  Implemented premultiplied alpha support: png_set_alpha_mode API
3365
3366Version 1.5.3beta07 [May 11, 2011]
3367  Added expand_16 support to the high level interface.
3368  Added named value and 'flag' gamma support to png_set_gamma.  Made a minor
3369    change from the previous (unreleased) ABI/API to hide the exact value used
3370    for Macs - it's not a good idea to embed this in the ABI!
3371  Moved macro definitions for PNG_HAVE_IHDR, PNG_HAVE_PLTE, and PNG_AFTER_IDAT
3372    from pngpriv.h to png.h because they must be visible to applications
3373    that call png_set_unknown_chunks().
3374  Check for up->location !PNG_AFTER_IDAT when writing unknown chunks
3375    before IDAT.
3376
3377Version 1.5.3beta08 [May 16, 2011]
3378  Improved "pngvalid --speed" to exclude more of pngvalid from the time.
3379  Documented png_set_alpha_mode(), other changes in libpng.3/libpng-manual.txt
3380  The cHRM chunk now sets the defaults for png_set_rgb_to_gray() (when negative
3381    parameters are supplied by the caller), while in the absence of cHRM
3382    sRGB/Rec 709 values are still used.  This introduced a divide-by-zero
3383    bug in png_handle_cHRM().
3384  The bKGD chunk no longer overwrites the background value set by
3385    png_set_background(), allowing the latter to be used before the file
3386    header is read. It never performed any useful function to override
3387    the default anyway.
3388  Added memory overwrite and palette image checks to pngvalid.c
3389    Previously palette image code was poorly checked. Since the transformation
3390    code has a special palette path in most cases this was a severe weakness.
3391  Minor cleanup and some extra checking in pngrutil.c and pngrtran.c. When
3392    expanding an indexed image, always expand to RGBA if transparency is
3393    present.
3394
3395Version 1.5.3beta09 [May 17, 2011]
3396  Reversed earlier 1.5.3 change of transformation order; move png_expand_16
3397    back where it was.  The change doesn't work because it requires 16-bit
3398    gamma tables when the code only generates 8-bit ones.  This fails
3399    silently; the libpng code just doesn't do any gamma correction.  Moving
3400    the tests back leaves the old, inaccurate, 8-bit gamma calculations, but
3401    these are clearly better than none!
3402
3403Version 1.5.3beta10 [May 20, 2011]
3404
3405  png_set_background() and png_expand_16() did not work together correctly.
3406    This problem is present in 1.5.2; if png_set_background is called with
3407    need_expand false and the matching 16 bit color libpng erroneously just
3408    treats it as an 8-bit color because of where png_do_expand_16 is in the
3409    transform list.  This simple fix reduces the supplied colour to 8-bits,
3410    so it gets smashed, but this is better than the current behavior.
3411  Added tests for expand16, more fixes for palette image tests to pngvalid.
3412    Corrects the code for palette image tests and disables attempts to
3413    validate palette colors.
3414
3415Version 1.5.3rc01 [June 3, 2011]
3416  No changes.
3417
3418Version 1.5.3rc02 [June 8, 2011]
3419  Fixed uninitialized memory read in png_format_buffer() (Bug report by
3420    Frank Busse, CVE-2011-2501, related to CVE-2004-0421).
3421
3422Version 1.5.3beta11 [June 11, 2011]
3423  Fixed png_handle_sCAL which is broken in 1.5. This fixes CVE 2011-2692.
3424  Added sCAL to pngtest.png
3425  Revised documentation about png_set_user_limits() to say that it also affects
3426    png writing.
3427  Revised handling of png_set_user_limits() so that it can increase the
3428    limit beyond the PNG_USER_WIDTH|HEIGHT_MAX; previously it could only
3429    reduce it.
3430  Make the 16-to-8 scaling accurate. Dividing by 256 with no rounding is
3431    wrong (high by one) 25% of the time. Dividing by 257 with rounding is
3432    wrong in 128 out of 65536 cases. Getting the right answer all the time
3433    without division is easy.
3434  Added "_SUPPORTED" to the PNG_WRITE_CUSTOMIZE_ZTXT_COMPRESSION macro.
3435  Added projects/owatcom, an IDE project for OpenWatcom to replace
3436    scripts/makefile.watcom.  This project works with OpenWatcom 1.9. The
3437    IDE autogenerates appropriate makefiles (libpng.mk) for batch processing.
3438    The project is configurable, unlike the Visual Studio project, so long
3439    as the developer has an awk.
3440  Changed png_set_gAMA to limit the gamma value range so that the inverse
3441    of the stored value cannot overflow the fixed point representation,
3442    and changed other things OpenWatcom warns about.
3443  Revised pngvalid.c to test PNG_ALPHA_MODE_SUPPORTED correctly. This allows
3444    pngvalid to build when ALPHA_MODE is not supported, which is required if
3445    it is to build on libpng 1.4.
3446  Removed string/memory macros that are no longer used and are not
3447    necessarily fully supportable, particularly png_strncpy and png_snprintf.
3448  Added log option to pngvalid.c and attempted to improve gamma messages.
3449
3450Version 1.5.3 [omitted]
3451  People found the presence of a beta release following an rc release
3452    to be confusing; therefore we bump the version to libpng-1.5.4beta01
3453    and there will be no libpng-1.5.3 release.
3454
3455Version 1.5.4beta01 [June 14, 2011]
3456  Made it possible to undefine PNG_READ_16_TO_8_ACCURATE_SCALE_SUPPORTED
3457    to get the same (inaccurate) output as libpng-1.5.2 and earlier.
3458  Moved definitions of PNG_HAVE_IHDR, PNG_AFTER_IDAT, and PNG_HAVE_PLTE
3459    outside of an unknown-chunk block in png.h because they are also
3460    needed for other uses.
3461
3462Version 1.5.4beta02 [June 14, 2011]
3463  Fixed and clarified LEGACY 16-to-8 scaling code.
3464  Added png_set_chop_16() API, to match inaccurate results from previous
3465    libpng versions.
3466  Removed the ACCURATE and LEGACY options (they are no longer useable)
3467  Use the old scaling method for background if png_set_chop_16() was
3468    called.
3469  Made png_set_chop_16() API removeable by disabling PNG_CHOP_16_TO_8_SUPPORTED
3470
3471Version 1.5.4beta03 [June 15, 2011]
3472  Fixed a problem in png_do_expand_palette() exposed by optimization in
3473    1.5.3beta06
3474  Also removed a spurious and confusing "trans" member ("trans") from png_info.
3475  The palette expand optimization prevented expansion to an intermediate RGBA
3476    form if tRNS was present but alpha was marked to be stripped; this exposed
3477    a check for tRNS in png_do_expand_palette() which is inconsistent with the
3478    code elsewhere in libpng.
3479  Correction to the expand_16 code; removed extra instance of
3480    png_set_scale_16_to_8 from pngpriv.h
3481
3482Version 1.5.4beta04 [June 16, 2011]
3483  Added a missing "#ifdef PNG_READ_BACKGROUND_SUPPORTED/#endif" in pngrtran.c
3484  Added PNG_TRANSFORM_CHOP_16 to the high-level read transforms.
3485  Made PNG_READ_16_TO_8_ACCURATE_SCALE configurable again.  If this is
3486    not enabled, png_set_strip_16() and png_do_scale_16_to_8() aren't built.
3487  Revised contrib/visupng, gregbook, and pngminim to demonstrate chop_16_to_8
3488
3489Version 1.5.4beta05 [June 16, 2011]
3490  Renamed png_set_strip_16() to png_set_scale_16() and renamed
3491    png_set_chop_16() to png_set_strip(16) in an attempt to minimize the
3492    behavior changes between libpng14 and libpng15.
3493
3494Version 1.5.4beta06 [June 18, 2011]
3495  Fixed new bug that was causing both strip_16 and scale_16 to be applied.
3496
3497Version 1.5.4beta07 [June 19, 2011]
3498  Fixed pngvalid, simplified macros, added checking for 0 in sCAL.
3499    The ACCURATE scale macro is no longer defined in 1.5 - call the
3500    png_scale_16_to_8 API.  Made sure that PNG_READ_16_TO_8 is still defined
3501    if the png_strip_16_to_8 API is present.  png_check_fp_number now
3502    maintains some state so that positive, negative and zero values are
3503    identified.  sCAL uses these to be strictly spec conformant.
3504
3505Version 1.5.4beta08 [June 23, 2011]
3506  Fixed pngvalid if ACCURATE_SCALE is defined.
3507  Updated scripts/pnglibconf.h.prebuilt.
3508
3509Version 1.5.4rc01 [June 30, 2011]
3510  Define PNG_ALLOCATED to "restrict" only if MSC_VER >= 1400.
3511
3512Version 1.5.4 [July 7, 2011]
3513  No changes.
3514
3515Version 1.5.5beta01 [July 13, 2011]
3516  Fixed some typos and made other minor changes in the manual.
3517  Updated contrib/pngminus/makefile.std (Samuli Souminen)
3518
3519Version 1.5.5beta02 [July 14, 2011]
3520  Revised Makefile.am and Makefile.in to look in the right directory for
3521    pnglibconf.h.prebuilt
3522
3523Version 1.5.5beta03 [July 27, 2011]
3524  Enabled compilation with g++ compiler.  This compiler does not recognize
3525    the file extension, so it always compiles with C++ rules.  Made minor
3526    changes to pngrutil.c to cast results where C++ expects it but C does not.
3527  Minor editing of libpng.3 and libpng-manual.txt.
3528
3529Version 1.5.5beta04 [July 29, 2011]
3530  Revised CMakeLists.txt (Clifford Yapp)
3531  Updated commentary about the png_rgb_to_gray() default coefficients
3532    in the manual and in pngrtran.c
3533
3534Version 1.5.5beta05 [August 17, 2011]
3535  Prevent unexpected API exports from non-libpng DLLs on Windows.  The "_DLL"
3536    is removed from the test of whether a DLL is being built (this erroneously
3537    caused the libpng APIs to be marked as DLL exports in static builds under
3538    Microsoft Visual Studio).  Almost all of the libpng building configuration
3539    is moved from pngconf.h to pngpriv.h, but PNG_DLL_EXPORT remains in
3540    pngconf.h, though, so that it is colocated with the import definition (it
3541    is no longer used anywhere in the installed headers).  The VStudio project
3542    definitions have been cleaned up: "_USRDLL" has been removed from the
3543    static library builds (this was incorrect), and PNG_USE_DLL has been added
3544    to pngvalid to test the functionality (pngtest does not supply it,
3545    deliberately).  The spurious "_EXPORTS" has been removed from the
3546    libpng build (all these errors were a result of copy/paste between project
3547    configurations.)
3548  Added new types and internal functions for CIE RGB end point handling to
3549    pngpriv.h (functions yet to be implemented).
3550
3551Version 1.5.5beta06 [August 26, 2011]
3552  Ensure the CMAKE_LIBRARY_OUTPUT_DIRECTORY is set in CMakeLists.txt
3553    (Clifford Yap)
3554  Fixes to rgb_to_gray and cHRM XYZ APIs (John Bowler):
3555    The rgb_to_gray code had errors when combined with gamma correction.
3556    Some pixels were treated as true grey when they weren't and such pixels
3557    and true grey ones were not gamma corrected (the original value of the
3558    red component was used instead).  APIs to get and set cHRM using color
3559    space end points have been added and the rgb_to_gray code that defaults
3560    based on cHRM, and the divide-by-zero bug in png_handle_cHRM (CERT
3561    VU#477046, CVE-2011-3328, introduced in 1.5.4) have been corrected.
3562  A considerable number of tests has been added to pngvalid for the
3563    rgb_to_gray transform.
3564  Arithmetic errors in rgb_to_gray whereby the calculated gray value was
3565    truncated to the bit depth rather than rounded have been fixed except in
3566    the 8-bit non-gamma-corrected case (where consistency seems more important
3567    than correctness.)  The code still has considerable inaccuracies in the
3568    8-bit case because 8-bit linear arithmetic is used.
3569
3570Version 1.5.5beta07 [September 7, 2011]
3571  Added "$(ARCH)" option to makefile.darwin
3572  Added SunOS support to configure.ac and Makefile.am
3573  Changed png_chunk_benign_error() to png_warning() in png.c, in
3574    png_XYZ_from_xy_checked().
3575
3576Version 1.5.5beta08 [September 10, 2011]
3577  Fixed 64-bit compilation errors (gcc). The errors fixed relate
3578    to conditions where types that are 32 bits in the GCC 32-bit
3579    world (uLong and png_size_t) become 64 bits in the 64-bit
3580    world.  This produces potential truncation errors which the
3581    compiler correctly flags.
3582  Relocated new HAVE_SOLARIS_LD definition in configure.ac
3583  Constant changes for 64-bit compatibility (removal of L suffixes). The
3584    16-bit cases still use "L" as we don't have a 16-bit test system.
3585
3586Version 1.5.5rc01 [September 15, 2011]
3587  Removed "L" suffixes in pngpriv.h
3588
3589Version 1.5.5 [September 22, 2011]
3590  No changes.
3591
3592Version 1.5.6beta01 [September 22, 2011]
3593  Fixed some 64-bit type conversion warnings in pngrtran.c
3594  Moved row_info from png_struct to a local variable.
3595  The various interlace mask arrays have been made into arrays of
3596    bytes and made PNG_CONST and static (previously some arrays were
3597    marked PNG_CONST and some weren't).
3598  Additional checks have been added to the transform code to validate the
3599    pixel depths after the transforms on both read and write.
3600  Removed some redundant code from pngwrite.c, in png_destroy_write_struct().
3601  Changed chunk reading/writing code to use png_uint_32 instead of png_byte[4].
3602    This removes the need to allocate temporary strings for chunk names on
3603    the stack in the read/write code.  Unknown chunk handling still uses the
3604    string form because this is exposed in the API.
3605
3606Version 1.5.6beta02 [September 26, 2011]
3607  Added a note in the manual the png_read_update_info() must be called only
3608    once with a particular info_ptr.
3609  Fixed a typo in the definition of the new PNG_STRING_FROM_CHUNK(s,c) macro.
3610
3611Version 1.5.6beta03 [September 28, 2011]
3612  Revised test-pngtest.sh to report FAIL when pngtest fails.
3613  Added "--strict" option to pngtest, to report FAIL when the failure is
3614    only because the resulting valid files are different.
3615  Revised CMakeLists.txt to work with mingw and removed some material from
3616    CMakeLists.txt that is no longer useful in libpng-1.5.
3617
3618Version 1.5.6beta04 [October 5, 2011]
3619  Fixed typo in Makefile.in and Makefile.am ("-M Wl" should be "-M -Wl")."
3620
3621Version 1.5.6beta05 [October 12, 2011]
3622  Speed up png_combine_row() for interlaced images. This reduces the generality
3623    of the code, allowing it to be optimized for Adam7 interlace.  The masks
3624    passed to png_combine_row() are now generated internally, avoiding
3625    some code duplication and localizing the interlace handling somewhat.
3626  Align png_struct::row_buf - previously it was always unaligned, caused by
3627    a bug in the code that attempted to align it; the code needs to subtract
3628    one from the pointer to take account of the filter byte prepended to
3629    each row.
3630  Optimized png_combine_row() when rows are aligned. This gains a small
3631    percentage for 16-bit and 32-bit pixels in the typical case where the
3632    output row buffers are appropriately aligned. The optimization was not
3633    previously possible because the png_struct buffer was always misaligned.
3634  Fixed bug in png_write_chunk_header() debug print, introduced in 1.5.6beta01.
3635
3636Version 1.5.6beta06 [October 17, 2011]
3637  Removed two redundant tests for unitialized row.
3638  Fixed a relatively harmless memory overwrite in compressed text writing
3639    with a 1 byte zlib buffer.
3640  Add ability to call png_read_update_info multiple times to pngvalid.c.
3641  Fixes for multiple calls to png_read_update_info. These fixes attend to
3642    most of the errors revealed in pngvalid, however doing the gamma work
3643    twice results in inaccuracies that can't be easily fixed.  There is now
3644    a warning in the code if this is going to happen.
3645  Turned on multiple png_read_update_info in pngvalid transform tests.
3646  Prevent libpng from overwriting unused bits at the end of the image when
3647    it is not byte aligned, while reading. Prior to libpng-1.5.6 libpng would
3648    overwrite the partial byte at the end of each row if the row width was not
3649    an exact multiple of 8 bits and the image is not interlaced.
3650
3651Version 1.5.6beta07 [October 21, 2011]
3652  Made png_ptr->prev_row an aligned pointer into png_ptr->big_prev_row
3653    (Mans Rullgard).
3654
3655Version 1.5.6rc01 [October 26, 2011]
3656  Changed misleading "Missing PLTE before cHRM" warning to "Out of place cHRM"
3657
3658Version 1.5.6rc02 [October 27, 2011]
3659  Added LSR() macro to defend against buggy compilers that evaluate non-taken
3660    code branches and complain about out-of-range shifts.
3661
3662Version 1.5.6rc03 [October 28, 2011]
3663  Renamed the LSR() macro to PNG_LSR() and added PNG_LSL() macro.
3664  Fixed compiler warnings with Intel and MSYS compilers. The logical shift
3665    fix for Microsoft Visual C is required by other compilers, so this
3666    enables that fix for all compilers when using compile-time constants.
3667    Under MSYS 'byte' is a name declared in a system header file, so we
3668    changed the name of a local variable to avoid the warnings that result.
3669  Added #define PNG_ALIGN_TYPE PNG_ALIGN_NONE to contrib/pngminim/*/pngusr.h
3670
3671Version 1.5.6 [November 3, 2011]
3672  No changes.
3673
3674Version 1.5.7beta01 [November 4, 2011]
3675  Added support for ARM processor, when decoding all PNG up-filtered rows
3676    and any other-filtered rows with 3 or 4 bytes per pixel (Mans Rullgard).
3677  Fixed bug in pngvalid on early allocation failure; fixed type cast in
3678    pngmem.c; pngvalid would attempt to call png_error() if the allocation
3679    of a png_struct or png_info failed. This would probably have led to a
3680    crash.  The pngmem.c implementation of png_malloc() included a cast
3681    to png_size_t which would fail on large allocations on 16-bit systems.
3682  Fix for the preprocessor of the Intel C compiler. The preprocessor
3683    splits adjacent @ signs with a space; this changes the concatentation
3684    token from @-@-@ to PNG_JOIN; that should work with all compiler
3685    preprocessors.
3686  Paeth filter speed improvements from work by Siarhei Siamashka. This
3687    changes the 'Paeth' reconstruction function to improve the GCC code
3688    generation on x86. The changes are only part of the suggested ones;
3689    just the changes that definitely improve speed and remain simple.
3690    The changes also slightly increase the clarity of the code.
3691
3692Version 1.5.7beta02 [November 11, 2011]
3693  Check compression_type parameter in png_get_iCCP and remove spurious
3694    casts. The compression_type parameter is always assigned to, so must
3695    be non-NULL. The cast of the profile length potentially truncated the
3696    value unnecessarily on a 16-bit int system, so the cast of the (byte)
3697    compression type to (int) is specified by ANSI-C anyway.
3698  Fixed FP division by zero in pngvalid.c; the 'test_pixel' code left
3699    the sBIT fields in the test pixel as 0, which resulted in a floating
3700    point division by zero which was irrelevant but causes systems where
3701    FP exceptions cause a crash. Added code to pngvalid to turn on FP
3702    exceptions if the appropriate glibc support is there to ensure this is
3703    tested in the future.
3704  Updated scripts/pnglibconf.mak and scripts/makefile.std to handle the
3705    new PNG_JOIN macro.
3706  Added versioning to pnglibconf.h comments.
3707  Simplified read/write API initial version; basic read/write tested on
3708    a variety of images, limited documentation (in the header file.)
3709  Installed more accurate linear to sRGB conversion tables. The slightly
3710    modified tables reduce the number of 16-bit values that
3711    convert to an off-by-one 8-bit value.  The "makesRGB.c" code that was used
3712    to generate the tables is now in a contrib/sRGBtables sub-directory.
3713
3714Version 1.5.7beta03 [November 17, 2011]
3715  Removed PNG_CONST from the sRGB table declarations in pngpriv.h and png.c
3716  Added run-time detection of NEON support.
3717  Added contrib/libtests; includes simplified API test and timing test and
3718    a color conversion utility for rapid checking of failed 'pngstest' results.
3719  Multiple transform bug fixes plus a work-round for double gamma correction.
3720    libpng does not support more than one transform that requires linear data
3721    at once - if this is tried typically the results is double gamma
3722    correction. Since the simplified APIs can need rgb to gray combined with
3723    a compose operation it is necessary to do one of these outside the main
3724    libpng transform code. This check-in also contains fixes to various bugs
3725    in the simplified APIs themselves and to some bugs in compose and rgb to
3726    gray (on palette) itself.
3727  Fixes for C++ compilation using g++ When libpng source is compiled
3728    using g++. The compiler imposes C++ rules on the C source; thus it
3729    is desireable to make the source work with either C or C++ rules
3730    without throwing away useful error information.  This change adds
3731    png_voidcast to allow C semantic (void*) cases or the corresponding
3732    C++ static_cast operation, as appropriate.
3733  Added --noexecstack to assembler file compilation. GCC does not set
3734    this on assembler compilation, even though it does on C compilation.
3735    This creates security issues if assembler code is enabled; the
3736    work-around is to set it by default in the flags for $(CCAS)
3737  Work around compilers that don't support declaration of const data. Some
3738    compilers fault 'extern const' data declarations (because the data is
3739    not initialized); this turns on const-ness only for compilers where
3740    this is known to work.
3741
3742Version 1.5.7beta04 [November 17, 2011]
3743  Since the gcc driver does not recognize the --noexecstack flag, we must
3744    use the -Wa prefix to have it passed through to the assembler.
3745    Also removed a duplicate setting of this flag.
3746  Added files that were omitted from the libpng-1.5.7beta03 zip distribution.
3747
3748Version 1.5.7beta05 [November 25, 2011]
3749  Removed "zTXt" from warning in generic chunk decompression function.
3750  Validate time settings passed to pngset() and png_convert_to_rfc1123()
3751    (Frank Busse).
3752  Added MINGW support to CMakeLists.txt
3753  Reject invalid compression flag or method when reading the iTXt chunk.
3754  Backed out 'simplified' API changes. The API seems too complex and there
3755    is a lack of consensus or enthusiasm for the proposals.  The API also
3756    reveals significant bugs inside libpng (double gamma correction and the
3757    known bug of being unable to retrieve a corrected palette). It seems
3758    better to wait until the bugs, at least, are corrected.
3759  Moved pngvalid.c into contrib/libtests
3760  Rebuilt Makefile.in, configure, etc., with autoconf-2.68
3761
3762Version 1.5.7rc01 [December 1, 2011]
3763  Replaced an "#if" with "#ifdef" in pngrtran.c
3764  Revised #if PNG_DO_BC block in png.c (use #ifdef and add #else)
3765
3766Version 1.5.7rc02 [December 5, 2011]
3767  Revised project files and contrib/pngvalid/pngvalid.c to account for
3768    the relocation of pngvalid into contrib/libtests.
3769  Revised pngconf.h to use " __declspec(restrict)" only when MSC_VER >= 1400,
3770    as in libpng-1.5.4.
3771  Put CRLF line endings in the owatcom project files.
3772
3773Version 1.5.7rc03 [December 7, 2011]
3774  Updated CMakeLists.txt to account for the relocation of pngvalid.c
3775
3776Version 1.5.7 [December 15, 2011]
3777  Minor fixes to pngvalid.c for gcc 4.6.2 compatibility to remove warnings
3778    reported by earlier versions.
3779  Fixed minor memset/sizeof errors in pngvalid.c.
3780
3781Version 1.6.0beta01 [December 15, 2011]
3782  Removed machine-generated configure files from the GIT repository (they will
3783    continue to appear in the tarball distributions and in the libpng15 and
3784    earlier GIT branches).
3785  Restored the new 'simplified' API, which was started in libpng-1.5.7beta02
3786    but later deleted from libpng-1.5.7beta05.
3787  Added example programs for the new 'simplified' API.
3788  Added ANSI-C (C90) headers and require them, and take advantage of the
3789    change. Also fixed some of the projects/* and contrib/* files that needed
3790    updates for libpng16 and the move of pngvalid.c.
3791    With this change the required ANSI-C header files are assumed to exist: the
3792    implementation must provide float.h, limits.h, stdarg.h and stddef.h and
3793    libpng relies on limits.h and stddef.h existing and behaving as defined
3794    (the other two required headers aren't used).  Non-ANSI systems that don't
3795    have stddef.h or limits.h will have to provide an appropriate fake
3796    containing the relevant types and #defines.
3797  The use of FAR/far has been eliminated and the definition of png_alloc_size_t
3798    is now controlled by a flag so that 'small size_t' systems can select it
3799    if necessary.  Libpng 1.6 may not currently work on such systems -- it
3800    seems likely that it will ask 'malloc' for more than 65535 bytes with any
3801    image that has a sufficiently large row size (rather than simply failing
3802    to read such images).
3803  New tools directory containing tools used to generate libpng code.
3804  Fixed race conditions in parallel make builds. With higher degrees of
3805    parallelism during 'make' the use of the same temporary file names such
3806    as 'dfn*' can result in a race where a temporary file from one arm of the
3807    build is deleted or overwritten in another arm.  This changes the
3808    temporary files for suffix rules to always use $* and ensures that the
3809    non-suffix rules use unique file names.
3810
3811Version 1.6.0beta02 [December 21, 2011]
3812  Correct configure builds where build and source directories are separate.
3813    The include path of 'config.h' was erroneously made relative in pngvalid.c
3814    in libpng 1.5.7.
3815
3816Version 1.6.0beta03 [December 22, 2011]
3817  Start-up code size improvements, error handler flexibility. These changes
3818    alter how the tricky allocation of the initial png_struct and png_info
3819    structures are handled. png_info is now handled in pretty much the same
3820    way as everything else, except that the allocations handle NULL return
3821    silently.  png_struct is changed in a similar way on allocation and on
3822    deallocation a 'safety' error handler is put in place (which should never
3823    be required).  The error handler itself is changed to permit mismatches
3824    in the application and libpng error buffer size; however, this means a
3825    silent change to the API to return the jmp_buf if the size doesn't match
3826    the size from the libpng compilation; libpng now allocates the memory and
3827    this may fail.  Overall these changes result in slight code size
3828    reductions; however, this is a reduction in code that is always executed
3829    so is particularly valuable.  Overall on a 64-bit system the libpng DLL
3830    decreases in code size by 1733 bytes.  pngerror.o increases in size by
3831    about 465 bytes because of the new functionality.
3832  Added png_convert_to_rfc1123_buffer() and deprecated png_convert_to_rfc1123()
3833    to avoid including a spurious buffer in the png_struct.
3834
3835Version 1.6.0beta04 [December 30, 2011]
3836  Regenerated configure scripts with automake-1.11.2
3837  Eliminated png_info_destroy(). It is now used only in png.c and only calls
3838    one other internal function and memset().
3839  Enabled png_get_sCAL_fixed() if floating point APIs are enabled. Previously
3840    it was disabled whenever internal fixed point arithmetic was selected,
3841    which meant it didn't exist even on systems where FP was available but not
3842    preferred.
3843  Added pngvalid.c compile time checks for const APIs.
3844  Implemented 'restrict' for png_info and png_struct. Because of the way
3845    libpng works both png_info and png_struct are always accessed via a
3846    single pointer.  This means adding C99 'restrict' to the pointer gives
3847    the compiler some opportunity to optimize the code.  This change allows
3848    that.
3849  Moved AC_MSG_CHECKING([if libraries can be versioned]) later to the proper
3850    location in configure.ac (Gilles Espinasse).
3851  Changed png_memcpy to C assignment where appropriate. Changed all those
3852    uses of png_memcpy that were doing a simple assignment to assignments
3853    (all those cases where the thing being copied is a non-array C L-value).
3854  Added some error checking to png_set_*() routines.
3855  Removed the reference to the non-exported function png_memcpy() from
3856    example.c.
3857  Fixed the Visual C 64-bit build - it requires jmp_buf to be aligned, but
3858    it had become misaligned.
3859  Revised contrib/pngminus/pnm2png.c to avoid warnings when png_uint_32
3860    and unsigned long are of different sizes.
3861
3862Version 1.6.0beta05 [January 15, 2012]
3863  Updated manual with description of the simplified API (copied from png.h)
3864  Fix bug in pngerror.c: some long warnings were being improperly truncated
3865    (CVE-2011-3464, bug introduced in libpng-1.5.3beta05).
3866
3867Version 1.6.0beta06 [January 24, 2012]
3868  Added palette support to the simplified APIs. This commit
3869    changes some of the macro definitions in png.h, app code
3870    may need corresponding changes.
3871  Increased the formatted warning buffer to 192 bytes.
3872  Added color-map support to simplified API. This is an initial version for
3873    review; the documentation has not yet been updated.
3874  Fixed Min/GW uninstall to remove libpng.dll.a
3875
3876Version 1.6.0beta07 [January 28, 2012]
3877  Eliminated Intel icc/icl compiler warnings. The Intel (GCC derived)
3878    compiler issues slightly different warnings from those issued by the
3879    current vesions of GCC. This eliminates those warnings by
3880    adding/removing casts and small code rewrites.
3881  Updated configure.ac from autoupdate: added --enable-werror option.
3882    Also some layout regularization and removal of introduced tab characters
3883    (replaced with 3-character indentation).  Obsolete macros identified by
3884    autoupdate have been removed; the replacements are all in 2.59 so
3885    the pre-req hasn't been changed.  --enable-werror checks for support
3886    for -Werror (or the given argument) in the compiler.  This mimics the
3887    gcc configure option by allowing -Werror to be turned on safely; without
3888    the option the tests written in configure itself fail compilation because
3889    they cause compiler warnings.
3890  Rewrote autogen.sh to run autoreconf instead of running tools one-by-one.
3891  Conditionalize the install rules for MINGW and CYGWIN in CMakeLists.txt and
3892    set CMAKE_LIBRARY_OUTPUT_DIRECTORY to "lib" on all platforms (C. Yapp).
3893  Freeze libtool files in the 'scripts' directory. This version of autogen.sh
3894    attempts to dissuade people from running it when it is not, or should not,
3895    be necessary.  In fact, autogen.sh does not work when run in a libpng
3896    directory extracted from a tar distribution anymore. You must run it in
3897    a GIT clone instead.
3898  Added two images to contrib/pngsuite (1-bit and 2-bit transparent grayscale),
3899    and renamed three whose names were inconsistent with those in
3900    pngsuite/README.txt.
3901
3902Version 1.6.0beta08 [February 1, 2012]
3903  Fixed Image::colormap misalignment in pngstest.c
3904  Check libtool/libtoolize version number (2.4.2) in configure.ac
3905  Divide test-pngstest.sh into separate pngstest runs for basic and
3906    transparent images.
3907  Moved automake options to AM_INIT_AUTOMAKE in configure.ac
3908  Added color-tests, silent-rules (Not yet implemented in Makefile.am) and
3909    version checking to configure.ac
3910  Improved pngstest speed by not doing redundant tests and add const to
3911    the background parameter of png_image_finish_read. The --background
3912    option is now done automagically only when required, so that commandline
3913    option no longer exists.
3914  Cleaned up pngpriv.h to consistently declare all functions and data.
3915    Also eliminated PNG_CONST_DATA, which is apparently not needed but we
3916    can't be sure until it is gone.
3917  Added symbol prefixing that allows all the libpng external symbols
3918    to be prefixed (suggested by Reuben Hawkins).
3919  Updated "ftbb*.png" list in the owatcom and vstudio projects.
3920  Fixed 'prefix' builds on clean systems. The generation of pngprefix.h
3921    should not require itself.
3922  Updated INSTALL to explain that autogen.sh must be run in a GIT clone,
3923    not in a libpng directory extracted from a tar distribution.
3924
3925Version 1.6.0beta09 [February 1, 2012]
3926  Reverted the prebuilt configure files to libpng-1.6.0beta05 condition.
3927
3928Version 1.6.0beta10 [February 3, 2012]
3929  Added Z_SOLO for zlib-1.2.6+ and correct pngstest tests
3930  Updated list of test images in CMakeLists.txt
3931  Updated the prebuilt configure files to current condition.
3932  Revised INSTALL information about autogen.sh; it works in tar distributions.
3933
3934Version 1.6.0beta11 [February 16, 2012]
3935  Fix character count in pngstest command in projects/owatcom/pngstest.tgt
3936  Revised test-pngstest.sh to report PASS/FAIL for each image.
3937  Updated documentation about the simplified API.
3938  Corrected estimate of error in libpng png_set_rgb_to_gray API.  The API is
3939    extremely inaccurate for sRGB conversions because it uses an 8-bit
3940    intermediate linear value and it does not use the sRGB transform, so it
3941    suffers from the known instability in gamma transforms for values close
3942    to 0 (see Poynton).  The net result is that the calculation has a maximum
3943    error of 14.99/255; 0.5/255^(1/2.2).  pngstest now uses 15 for the
3944    permitted 8-bit error. This may still not be enough because of arithmetic
3945    error.
3946  Removed some unused arrays (with #ifdef) from png_read_push_finish_row().
3947  Fixed a memory overwrite bug in simplified read of RGB PNG with
3948    non-linear gamma Also bugs in the error checking in pngread.c and changed
3949    quite a lot of the checks in pngstest.c to be correct; either correctly
3950    written or not over-optimistic.  The pngstest changes are insufficient to
3951    allow all possible RGB transforms to be passed; pngstest cmppixel needs
3952    to be rewritten to make it clearer which errors it allows and then changed
3953    to permit known inaccuracies.
3954  Removed tests for no-longer-used *_EMPTY_PLTE_SUPPORTED from pngstruct.h
3955  Fixed fixed/float API export conditionals. 1) If FIXED_POINT or
3956    FLOATING_POINT options were switched off, png.h ended up with lone ';'
3957    characters.  This is not valid ANSI-C outside a function.  The ';'
3958    characters have been moved inside the definition of PNG_FP_EXPORT and
3959    PNG_FIXED_EXPORT. 2) If either option was switched off, the declaration
3960    of the corresponding functions were completely omitted, even though some
3961    of them are still used internally.  The result is still valid, but
3962    produces warnings from gcc with some warning options (including -Wall). The
3963    fix is to cause png.h to declare the functions with PNG_INTERNAL_FUNCTION
3964    when png.h is included from pngpriv.h.
3965  Check for invalid palette index while reading paletted PNG.  When one is
3966    found, issue a warning and increase png_ptr->num_palette accordingly.
3967    Apps are responsible for checking to see if that happened.
3968
3969Version 1.6.0beta12 [February 18, 2012]
3970  Do not increase num_palette on invalid_index.
3971  Relocated check for invalid palette index to pngrtran.c, after unpacking
3972    the sub-8-bit pixels.
3973  Fixed CVE-2011-3026 buffer overrun bug.  This bug was introduced when
3974    iCCP chunk support was added at libpng-1.0.6. Deal more correctly with the
3975    test on iCCP chunk length. Also removed spurious casts that may hide
3976    problems on 16-bit systems.
3977
3978Version 1.6.0beta13 [February 24, 2012]
3979  Eliminated redundant png_push_read_tEXt|zTXt|iTXt|unknown code from
3980    pngpread.c and use the sequential png_handle_tEXt, etc., in pngrutil.c;
3981    now that png_ptr->buffer is inaccessible to applications, the special
3982    handling is no longer useful.
3983  Added PNG_SAFE_LIMITS feature to pnglibconf.dfa, pngpriv.h, and new
3984    pngusr.dfa to reset the user limits to safe ones if PNG_SAFE_LIMITS is
3985    defined.  To enable, use "CPPFLAGS=-DPNG_SAFE_LIMITS_SUPPORTED=1" on the
3986    configure command or put #define PNG_SAFE_LIMITS_SUPPORTED in
3987    pnglibconf.h.prebuilt and pnglibconf.h.
3988
3989Version 1.6.0beta14 [February 27, 2012]
3990  Added information about the new limits in the manual.
3991  Updated Makefile.in
3992
3993Version 1.6.0beta15 [March 2, 2012]
3994  Removed unused "current_text" members of png_struct and the png_free()
3995    of png_ptr->current_text from pngread.c
3996  Rewrote pngstest.c for substantial speed improvement.
3997  Fixed transparent pixel and 16-bit rgb tests in pngstest and removed a
3998    spurious check in pngwrite.c
3999  Added PNG_IMAGE_FLAG_FAST for the benefit of applications that store
4000    intermediate files, or intermediate in-memory data, while processing
4001    image data with the simplified API.  The option makes the files larger
4002    but faster to write and read.  pngstest now uses this by default; this
4003    can be disabled with the --slow option.
4004  Improved pngstest fine tuning of error numbers, new test file generator.
4005    The generator generates images that test the full range of sample values,
4006    allow the error numbers in pngstest to be tuned and checked.  makepng
4007    also allows generation of images with extra chunks, although this is
4008    still work-in-progress.
4009  Added check for invalid palette index while reading.
4010  Fixed some bugs in ICC profile writing. The code should now accept
4011    all potentially valid ICC profiles and reject obviously invalid ones.
4012    It now uses png_error() to do so rather than casually writing a PNG
4013    without the necessary color data.
4014  Removed whitespace from the end of lines in all source files and scripts.
4015
4016Version 1.6.0beta16 [March 6, 2012]
4017  Relocated palette-index checking function from pngrutil.c to pngtrans.c
4018  Added palette-index checking while writing.
4019  Changed png_inflate() and calling routines to avoid overflow problems.
4020    This is an intermediate check-in that solves the immediate problems and
4021    introduces one performance improvement (avoiding a copy via png_ptr->zbuf.)
4022    Further changes will be made to make ICC profile handling more secure.
4023  Fixed build warnings (MSVC, GCC, GCC v3). Cygwin GCC with default options
4024    declares 'index' as a global, causing a warning if it is used as a local
4025    variable.  GCC 64-bit warns about assigning a (size_t) (unsigned 64-bit)
4026    to an (int) (signed 32-bit).  MSVC, however, warns about using the
4027    unary '-' operator on an unsigned value (even though it is well defined
4028    by ANSI-C to be ~x+1).  The padding calculation was changed to use a
4029    different method.  Removed the tests on png_ptr->pass.
4030  Added contrib/libtests/tarith.c to test internal arithmetic functions from
4031    png.c. This is a libpng maintainer program used to validate changes to the
4032    internal arithmetic functions.
4033  Made read 'inflate' handling like write 'deflate' handling. The read
4034    code now claims and releases png_ptr->zstream, like the write code.
4035    The bug whereby the progressive reader failed to release the zstream
4036    is now fixed, all initialization is delayed, and the code checks for
4037    changed parameters on deflate rather than always calling
4038    deflatedEnd/deflateInit.
4039  Validate the zTXt strings in pngvalid.
4040  Added code to validate the windowBits value passed to deflateInit2().
4041    If the call to deflateInit2() is wrong a png_warning will be issued
4042    (in fact this is harmless, but the PNG data produced may be sub-optimal).
4043
4044Version 1.6.0beta17 [March 10, 2012]
4045  Fixed PNG_LIBPNG_BUILD_BASE_TYPE definition.
4046  Reject all iCCP chunks after the first, even if the first one is invalid.
4047  Deflate/inflate was reworked to move common zlib calls into single
4048    functions [rw]util.c.  A new shared keyword check routine was also added
4049    and the 'zbuf' is no longer allocated on progressive read.  It is now
4050    possible to call png_inflate() incrementally.  A warning is no longer
4051    issued if the language tag or translated keyword in the iTXt chunk
4052    has zero length.
4053  If benign errors are disabled use maximum window on ancilliary inflate.
4054    This works round a bug introduced in 1.5.4 where compressed ancillary
4055    chunks could end up with a too-small windowBits value in the deflate
4056    header.
4057
4058Version 1.6.0beta18 [March 16, 2012]
4059  Issue a png_benign_error() instead of png_warning() about bad palette index.
4060  In pngtest, treat benign errors as errors if "-strict" is present.
4061  Fixed an off-by-one error in the palette index checking function.
4062  Fixed a compiler warning under Cygwin (Windows-7, 32-bit system)
4063  Revised example.c to put text strings in a temporary character array
4064    instead of directly assigning string constants to png_textp members.
4065    This avoids compiler warnings when -Wwrite-strings is enabled.
4066  Added output flushing to aid debugging under Visual Studio. Unfortunately
4067    this is necessary because the VS2010 output window otherwise simply loses
4068    the error messages on error (they weren't flushed to the window before
4069    the process exited, apparently!)
4070  Added configuration support for benign errors and changed the read
4071    default. Also changed some warnings in the iCCP and sRGB handling
4072    from to benign errors. Configuration now makes read benign
4073    errors warnings and write benign errors to errors by default (thus
4074    changing the behavior on read).  The simplified API always forces
4075    read benign errors to warnings (regardless of the system default, unless
4076    this is disabled in which case the simplified API can't be built.)
4077
4078Version 1.6.0beta19 [March 18, 2012]
4079  Work around for duplicate row start calls; added warning messages.
4080    This turns on PNG_FLAG_DETECT_UNINITIALIZED to detect app code that
4081    fails to call one of the 'start' routines (not enabled in libpng-1.5
4082    because it is technically an API change, since it did normally work
4083    before.)  It also makes duplicate calls to png_read_start_row (an
4084    internal function called at the start of the image read) benign, as
4085    they were before changes to use png_inflate_claim. Somehow webkit is
4086    causing this to happen; this is probably a mis-feature in the zlib
4087    changes so this commit is only a work-round.
4088  Removed erroneous setting of DETECT_UNINITIALIZED and added more
4089    checks. The code now does a png_error if an attempt is made to do the
4090    row initialization twice; this is an application error and it has
4091    serious consequences because the transform data in png_struct is
4092    changed by each call.
4093  Added application error reporting and added chunk names to read
4094    benign errors; also added --strict to pngstest - not enabled
4095    yet because a warning is produced.
4096  Avoid the double gamma correction warning in the simplified API.
4097    This allows the --strict option to pass in the pngstest checks
4098
4099Version 1.6.0beta20 [March 29, 2012]
4100  Changed chunk handler warnings into benign errors, incrementally load iCCP
4101  Added checksum-icc.c to contrib/tools
4102  Prevent PNG_EXPAND+PNG_SHIFT doing the shift twice.
4103  Recognize known sRGB ICC profiles while reading; prefer writing the
4104    iCCP profile over writing the sRGB chunk, controlled by the
4105    PNG_sRGB_PROFILE_CHECKS option.
4106  Revised png_set_text_2() to avoid potential memory corruption (fixes
4107    CVE-2011-3048, also known as CVE-2012-3425).
4108
4109Version 1.6.0beta21 [April 27, 2012]
4110  Revised scripts/makefile.darwin: use system zlib; remove quotes around
4111    architecture list; add missing ppc architecture; add architecture options
4112    to shared library link; don't try to create a shared lib based on missing
4113    RELEASE variable.
4114  Enable png_set_check_for_invalid_index() for both read and write.
4115  Removed #ifdef PNG_HANDLE_AS_UNKNOWN_SUPPORTED in pngpriv.h around
4116    declaration of png_handle_unknown().
4117  Added -lssp_nonshared in a comment in scripts/makefile.freebsd
4118    and changed deprecated NOOBJ and NOPROFILE to NO_OBJ and NO_PROFILE.
4119
4120Version 1.6.0beta22 [May 23, 2012]
4121  Removed need for -Wno-cast-align with clang.  clang correctly warns on
4122    alignment increasing pointer casts when -Wcast-align is passed. This
4123    fixes the cases that clang warns about either by eliminating the
4124    casts from png_bytep to png_uint_16p (pngread.c), or, for pngrutil.c
4125    where the cast is previously verified or pngstest.c where it is OK, by
4126    introducing new png_aligncast macros to do the cast in a way that clang
4127    accepts.
4128
4129Version 1.6.0beta23 [June 6, 2012]
4130  Revised CMakeLists.txt to not attempt to make a symlink under mingw.
4131  Made fixes for new optimization warnings from gcc 4.7.0. The compiler
4132    performs an optimization which is safe; however it then warns about it.
4133    Changing the type of 'palette_number' in pngvalid.c removes the warning.
4134  Do not depend upon a GCC feature macro being available for use in generating
4135    the linker mapfile symbol prefix.
4136  Improved performance of new do_check_palette_indexes() function (only
4137    update the value when it actually increases, move test for whether
4138    the check is wanted out of the function.
4139
4140Version 1.6.0beta24 [June 7, 2012]
4141  Don't check palette indexes if num_palette is 0 (as it can be in MNG files).
4142
4143Version 1.6.0beta25 [June 16, 2012]
4144  Revised png_set_keep_unknown_chunks() so num_chunks < 0 means ignore all
4145    unknown chunks and all known chunks except for IHDR, PLTE, tRNS, IDAT,
4146    and IEND.  Previously it only meant ignore all unknown chunks, the
4147    same as num_chunks == 0. Revised png_image_skip_unused_chunks() to
4148    provide a list of chunks to be processed instead of a list of chunks to
4149    ignore.  Revised contrib/gregbook/readpng2.c accordingly.
4150
4151Version 1.6.0beta26 [July 10, 2012]
4152  Removed scripts/makefile.cegcc from the *.zip and *.7z distributions; it
4153    depends on configure, which is not included in those archives.
4154  Moved scripts/chkfmt to contrib/tools.
4155  Changed "a+w" to "u+w" in Makefile.in to fix CVE-2012-3386.
4156
4157Version 1.6.0beta27 [August 11, 2012]
4158  Do not compile PNG_DEPRECATED, PNG_ALLOC and PNG_PRIVATE when __GNUC__ < 3.
4159  Do not use __restrict when GNUC is <= 3.1
4160  Removed references to png_zalloc() and png_zfree() from the manual.
4161  Fixed configurations where floating point is completely disabled.  Because
4162    of the changes to support symbol prefixing PNG_INTERNAL_FUNCTION declares
4163    floating point APIs during libpng builds even if they are completely
4164    disabled. This requires the png floating point types (png_double*) to be
4165    declared even though the functions are never actually defined.  This
4166    change provides a dummy definition so that the declarations work, yet any
4167    implementation will fail to compile because of an incomplete type.
4168  Re-eliminated the use of strcpy() in pngtest.c.  An unncessary use of
4169    strcpy() was accidentally re-introduced in libpng16; this change replaces
4170    it with strncpy().
4171  Eliminated use of png_sizeof(); use sizeof() instead.
4172  Use a consistent style for (sizeof type) and (sizeof (array))
4173  Cleanup of png_set_filler().  This function does very different things on
4174    read and write.  In libpng 1.6 the two cases can be distinguished and
4175    considerable code cleanup, and extra error checking, is possible.  This
4176    makes calls on the write side that have no effect be ignored with a
4177    png_app_error(), which can be disabled in the app using
4178    png_set_benign_errors(), and removes the spurious use of usr_channels
4179    on the read side.
4180  Insist on autotools 1.12.1 for git builds because there are security issues
4181    with 1.12 and insisting on anything less would allow 1.12 to be used.
4182  Removed info_ptr->signature[8] from WRITE-only builds.
4183  Add some conditions for compiling png_fixed().  This is a small function
4184    but it requires "-lm" on some platforms.
4185  Cause pngtest --strict to fail on any warning from libpng (not just errors)
4186    and cause it not to fail at the comparison step if libpng lacks support
4187    for writing chunks that it reads from the input (currently only implemented
4188    for compressed text chunks).
4189  Make all three "make check" test programs work without READ or WRITE support.
4190    Now "make check" will succeed even if libpng is compiled with -DPNG_NO_READ
4191    or -DPNG_NO_WRITE.  The tests performed are reduced, but the basic reading
4192    and writing of a PNG file is always tested by one or more of the tests.
4193  Consistently use strlen(), memset(), memcpy(), and memcmp() instead of the
4194    png_strlen(), png_memset(), png_memcpy(), and png_memcmp() macros.
4195  Removed the png_sizeof(), png_strlen(), png_memset(), png_memcpy(), and
4196    png_memcmp() macros.
4197  Work around gcc 3.x and Microsoft Visual Studio 2010 complaints. Both object
4198    to the split initialization of num_chunks.
4199
4200Version 1.6.0beta28 [August 29, 2012]
4201  Unknown handling fixes and clean up. This adds more correct option
4202    control of the unknown handling, corrects the pre-existing bug where
4203    the per-chunk 'keep' setting is ignored and makes it possible to skip
4204    IDAT chunks in the sequential reader (broken in earlier 1.6 versions).
4205    There is a new test program, test-unknown.c, which is a work in progress
4206    (not currently part of the test suite).  Comments in the header files now
4207    explain how the unknown handling works.
4208  Allow fine grain control of unknown chunk APIs. This change allows
4209    png_set_keep_unknown_chunks() to be turned off if not required and causes
4210    both read and write to behave appropriately (on read this is only possible
4211    if the user callback is used to handle unknown chunks).  The change
4212    also removes the support for storing unknown chunks in the info_struct
4213    if the only unknown handling enabled is via the callback, allowing libpng
4214    to be configured with callback reading and none of the unnecessary code.
4215  Corrected fix for unknown handling in pngtest. This reinstates the
4216    libpng handling of unknown chunks other than vpAg and sTER (including
4217    unsafe-to-copy chunks which were dropped before) and eliminates the
4218    repositioning of vpAg and sTER in pngtest.png by changing pngtest.png
4219    (so the chunks are where libpng would put them).
4220  Added "tunknown" test and corrected a logic error in png_handle_unknown()
4221    when SAVE support is absent.  Moved the shell test scripts for
4222    contrib/libtests from the libpng top directory to contrib/libtests.
4223    png_handle_unknown() must always read or skip the chunk, if
4224    SAVE_UNKNOWN_CHUNKS is turned off *and* the application does not set
4225    a user callback an unknown chunk will not be read, leading to a read
4226    error, which was revealed by the "tunknown" test.
4227  Cleaned up and corrected ICC profile handling.
4228    contrib/libtests/makepng: corrected 'rgb' and 'gray' cases.  profile_error
4229    messages could be truncated; made a correct buffer size calculation and
4230    adjusted pngerror.c appropriately. png_icc_check_* checking improved;
4231    changed the functions to receive the correct color type of the PNG on read
4232    or write and check that it matches the color space of the profile (despite
4233    what the comments said before, there is danger in assuming the app will
4234    cope correctly with an RGB profile on a grayscale image and, since it
4235    violates the PNG spec, allowing it is certain to produce inconsistent
4236    app behavior and might even cause app crashes.) Check that profiles
4237    contain the tags needed to process the PNG (tags all required by the ICC
4238    spec). Removed unused PNG_STATIC from pngpriv.h.
4239
4240Version 1.6.0beta29 [September 4, 2012]
4241  Fixed the simplified API example programs to add the *colormap parameter
4242    to several of he API and improved the error message if the version field
4243    is not set.
4244  Added contrib/examples/* to the *.zip and *.7z distributions.
4245  Updated simplified API synopses and description of the png_image structure
4246    in the manual.
4247  Made makepng and pngtest produce identical PNGs, add "--relaxed" option
4248    to pngtest. The "--relaxed" option turns off the benign errors that are
4249    enabled by default in pre-RC builds. makepng can now write ICC profiles
4250    where the length has not been extended to a multiple of 4, and pngtest
4251    now intercepts all libpng errors, allowing the previously-introduced
4252    "--strict test" on no warnings to actually work.
4253  Improved ICC profile handling including cHRM chunk generation and fixed
4254    Cygwin+MSVC build errors. The ICC profile handling now includes more
4255    checking.  Several errors that caused rejection of the profile are now
4256    handled with a warning in such a way that the invalid profiles will be
4257    read by default in release (but not pre-RC) builds but will not be
4258    written by default.  The easy part of handling the cHRM chunk is written,
4259    where the ICC profile contains the required data.  The more difficult
4260    part plus guessing a gAMA value requires code to pass selected RGB values
4261    through the profile.
4262
4263Version 1.6.0beta30 [October 24, 2012]
4264  Changed ICC profile matrix/vector types to not depend on array type rules.
4265    By the ANSI-C standard the new types should be identical to the previous
4266    versions, and all known versions of gcc tested with the previous versions
4267    except for GCC-4.2.1 work with this version.  The change makes the ANSI-C
4268    rule that const applied to an array of elements applies instead to the
4269    elements in the array moot by explicitly applying const to the base
4270    elements of the png_icc_matrix and png_icc_vector types. The accidental
4271    (harmless) 'const' previously applied to the parameters of two of the
4272    functions have also been removed.
4273  Added a work around for GCC 4.2 optimization bug.
4274  Marked the broken (bad white point) original HP sRGB profiles correctly and
4275    correct comments.
4276  Added -DZ_SOLO to contrib/pngminim/*/makefile to work with zlib-1.2.7
4277  Use /MDd for vstudio debug builds. Also added pngunkown to the vstudio
4278    builds, fixed build errors and corrected a minor exit code error in
4279    pngvalid if the 'touch' file name is invalid.
4280  Add updated WARNING file to projects/vstudio from libpng 1.5/vstudio
4281  Fixed build when using #define PNG_NO_READ_GAMMA in png_do_compose() in
4282    pngrtran.c (Domani Hannes).
4283
4284Version 1.6.0beta31 [November 1, 2012]
4285  Undid the erroneous change to vstudio/pngvalid build in libpng-1.6.0beta30.
4286  Made pngvalid so that it will build outside the libpng source tree.
4287  Made builds -DPNG_NO_READ_GAMMA compile (the unit tests still fail).
4288  Made PNG_NO_READ_GAMMA switch off interfaces that depend on READ_GAMMA.
4289    Prior to 1.6.0 switching off READ_GAMMA did unpredictable things to the
4290    interfaces that use it (specifically, png_do_background in 1.4 would
4291    simply display composite for grayscale images but do composition
4292    with the incorrect arithmetic for color ones). In 1.6 the semantic
4293    of -DPNG_NO_READ_GAMMA is changed to simply disable any interface that
4294    depends on it; this obliges people who set it to consider whether they
4295    really want it off if they happen to use any of the interfaces in
4296    question (typically most users who disable it won't).
4297  Fixed GUIDs in projects/vstudio. Some were duplicated or missing,
4298    resulting in VS2010 having to update the files.
4299  Removed non-working ICC profile support code that was mostly added to
4300    libpng-1.6.0beta29 and beta30. There was too much code for too little
4301    gain; implementing full ICC color correction may be desireable but is left
4302    up to applications.
4303
4304Version 1.6.0beta32 [November 25, 2012]
4305  Fixed an intermittent SEGV in pngstest due to an uninitialized array element.
4306  Added the ability for contrib/libtests/makepng.c to make a PNG with just one
4307    color. This is useful for debugging pngstest color inaccuracy reports.
4308  Fixed error checking in the simplified write API (Olaf van der Spek)
4309  Made png_user_version_check() ok to use with libpng version 1.10.x and later.
4310
4311Version 1.6.0beta33 [December 15, 2012]
4312  Fixed typo in png.c (PNG_SET_CHUNK_MALLOC_MAX should be PNG_CHUNK_MALLOC_MAX)
4313    that causes the MALLOC_MAX limit not to work (John Bowler)
4314  Change png_warning() to png_app_error() in pngwrite.c and comment the
4315    fall-through condition.
4316  Change png_warning() to png_app_warning() in png_write_tRNS().
4317  Rearranged the ARM-NEON optimizations: Isolated the machine specific code
4318    to the hardware subdirectory and added comments to pngrutil.c so that
4319    implementors of other optimizations know what to do.
4320  Fixed cases of unquoted DESTDIR in Makefile.am
4321  Rebuilt Makefile.in, etc., with autoconf-2.69 and automake-1.12.5.
4322
4323Version 1.6.0beta34 [December 19, 2012]
4324  Cleaned up whitespace in the synopsis portion of the manpage "libpng.3"
4325  Disassembled the version number in scripts/options.awk (necessary for
4326    building on SunOs).
4327
4328Version 1.6.0beta35 [December 23, 2012]
4329  Made default Zlib compression settings be configurable. This adds #defines to
4330    pnglibconf.h to control the defaults.
4331  Fixed Windows build issues, enabled ARM compilation. Various warnings issued
4332    by earlier versions of GCC fixed for Cygwin and Min/GW (which both use old
4333    GCCs.) ARM support is enabled by default in zlib.props (unsupported by
4334    Microsoft) and ARM compilation is made possible by deleting the check for
4335    x86. The test programs cannot be run because they are not signed.
4336
4337Version 1.6.0beta36 [January 2, 2013]
4338  Discontinued distributing libpng-1.x.x.tar.bz2.
4339  Discontinued distributing libpng-1.7.0-1.6.0-diff.txt and similar.
4340  Rebuilt configure with autoconf-2.69 (inadvertently not done in beta33)
4341  Fixed 'make distcheck' on SUN OS - libpng.so was not being removed
4342
4343Version 1.6.0beta37 [January 10, 2013]
4344  Fixed conceivable but difficult to repro overflow. Also added two test
4345    programs to generate and test a PNG which should have the problem.
4346
4347Version 1.6.0beta39 [January 19, 2013]
4348  Again corrected attempt at overflow detection in png_set_unknown_chunks().
4349  Added overflow detection in png_set_sPLT() and png_set_text_2().
4350
4351Version 1.6.0beta40 [January 20, 2013]
4352  Use consistent handling of overflows in text, sPLT and unknown png_set_* APIs
4353
4354Version 1.6.0rc01 [January 26, 2013]
4355  No changes.
4356
4357Version 1.6.0rc02 [February 4, 2013]
4358  Added png_get_palette_max() function.
4359
4360Version 1.6.0rc03 [February 5, 2013]
4361  Fixed the png_get_palette_max API.
4362
4363Version 1.6.0rc04 [February 7, 2013]
4364  Turn serial tests back on (recently turned off by autotools upgrade).
4365
4366Version 1.6.0rc05 [February 8, 2013]
4367  Update manual about png_get_palette_max().
4368
4369Version 1.6.0rc06 [February 9, 2013]
4370  Fixed missing dependency in --prefix builds The intermediate
4371    internal 'prefix.h' file can only be generated correctly after
4372    pnglibconf.h, however the dependency was not in Makefile.am.  The
4373    symptoms are unpredictable depending on the order make chooses to
4374    build pngprefix.h and pnglibconf.h, often the error goes unnoticed
4375    because there is a system pnglibconf.h to use instead.
4376
4377Version 1.6.0rc07 [February 10, 2013]
4378  Enclosed the new png_get_palette_max in #ifdef PNG_GET_PALETTE_MAX_SUPPORTED
4379    block, and revised pnglibconf.h and pnglibconf.h.prebuilt accordingly.
4380
4381Version 1.6.0rc08 [February 10, 2013]
4382  Fix typo in png.h #ifdef
4383
4384Version 1.6.0 [February 14, 2013]
4385  No changes.
4386
4387Version 1.6.1beta01 [February 16, 2013]
4388  Made symbol prefixing work with the ARM neon optimizations. Also allow
4389    pngpriv.h to be included for preprocessor definitions only, so it can
4390    be used in non-C/C++ files. Back ported from libpng 1.7.
4391  Made sRGB check numbers consistent.
4392  Ported libpng 1.5 options.awk/dfn file handling to 1.6, fixed one bug.
4393  Removed cc -E workround, corrected png_get_palette_max API Tested on
4394    SUN OS cc 5.9, which demonstrates the tokenization problem previously
4395    avoided by using /lib/cpp.  Since all .dfn output is now protected in
4396    double quotes unless it is to be macro substituted the fix should
4397    work everywhere.
4398  Enabled parallel tests - back ported from libpng-1.7.
4399  scripts/pnglibconf.dfa formatting improvements back ported from libpng17.
4400  Fixed a race condition in the creation of the build 'scripts' directory
4401    while building with a parallel make.
4402  Use approved/supported Android method to check for NEON, use Linux/POSIX
4403    1003.1 API to check /proc/self/auxv avoiding buffer allocation and other
4404    library calls (ported from libpng15).
4405
4406Version 1.6.1beta02 [February 19, 2013]
4407  Use parentheses more consistently in "#if defined(MACRO)" tests.
4408  Folded long lines.
4409  Reenabled code to allow zero length PLTE chunks for MNG.
4410
4411Version 1.6.1beta03 [February 22, 2013]
4412  Fixed ALIGNED_MEMORY support.
4413  Allow run-time ARM NEON checking to be disabled. A new configure option:
4414    --enable-arm-neon=always will stop the run-time checks. New checks
4415    within arm/arm_init.c will cause the code not to be compiled unless
4416    __ARM_NEON__ is set. This should make it fail safe (if someone asks
4417    for it on then the build will fail if it can't be done.)
4418  Updated the INSTALL document.
4419
4420Version 1.6.1beta04 [February 27, 2013]
4421  Revised INSTALL to recommend using CPPFLAGS instead of INCLUDES.
4422  Revised scripts/makefile.freebsd to respect ZLIBLIB and ZLIBINC.
4423  Revised scripts/dfn.awk to work with the buggy MSYS awk that has trouble
4424    with CRLF line endings.
4425
4426Version 1.6.1beta05 [March 1, 2013]
4427  Avoid a possible memory leak in contrib/gregbook/readpng.c
4428
4429Version 1.6.1beta06 [March 4, 2013]
4430  Better documentation of unknown handling API interactions.
4431  Corrected Android builds and corrected libpng.vers with symbol
4432    prefixing This adds an API to set optimization options externally,
4433    providing an alternative and general solution for the non-portable
4434    run-time tests used by the ARM Neon code.  It also makes those tests
4435    compile and link on Android.
4436  The order of settings vs options in pnglibconf.h is reversed to allow
4437    settings to depend on options and options can now set (or override) the
4438    defaults for settings.
4439
4440Version 1.6.1beta07 [March 7, 2013]
4441  Corrected simplified API default gamma for color-mapped output, added
4442    a flag to change default. In 1.6.0 when the simplified API was used
4443    to produce color-mapped output from an input image with no gamma
4444    information the gamma assumed for the input could be different from
4445    that assumed for non-color-mapped output.  In particular 16-bit depth
4446    input files were assumed to be sRGB encoded, whereas in the 'direct'
4447    case they were assumed to have linear data.  This was an error.  The
4448    fix makes the simplified API treat all input files the same way and
4449    adds a new flag to the png_image::flags member to allow the
4450    application/user to specify that 16-bit files contain sRGB data
4451    rather than the default linear.
4452  Fixed bugs in the pngpixel and makepng test programs.
4453
4454Version 1.6.1beta08 [March 7, 2013]
4455  Fixed CMakelists.txt to allow building a single variant of the library
4456    (Claudio Bley):
4457  Introduced a PNG_LIB_TARGETS variable that lists all activated library
4458    targets.  It is an error if this variable ends up empty, ie. you have
4459    to build at least one library variant.
4460  Made the *_COPY targets only depend on library targets actually being build.
4461  Use PNG_LIB_TARGETS to unify a code path.
4462  Changed the CREATE_SYMLINK macro to expect the full path to a file as the
4463    first argument. When symlinking the filename component of that path is
4464    determined and used as the link target.
4465  Use copy_if_different in the CREATE_SYMLINK macro.
4466
4467Version 1.6.1beta09 [March 13, 2013]
4468  Eliminated two warnings from the Intel C compiler. The warnings are
4469    technically valid, although a reasonable treatment of division would
4470    show it to be incorrect.
4471
4472Version 1.6.1rc01 [March 21, 2013]
4473  No changes.
4474
4475Version 1.6.1 [March 28, 2013]
4476  No changes.
4477
4478Version 1.6.2beta01 [April 14, 2013]
4479  Updated documentation of 1.5.x to 1.6.x changes in iCCP chunk handling.
4480  Fixed incorrect warning of excess deflate data. End condition - the
4481    warning would be produced if the end of the deflate stream wasn't read
4482    in the last row.  The warning is harmless.
4483  Corrected the test on user transform changes on read. It was in the
4484    png_set of the transform function, but that doesn't matter unless the
4485    transform function changes the rowbuf size, and that is only valid if
4486    transform_info is called.
4487  Corrected a misplaced closing bracket in contrib/libtests/pngvalid.c
4488    (Flavio Medeiros).
4489  Corrected length written to uncompressed iTXt chunks (Samuli Suominen).
4490    Bug was introduced in libpng-1.6.0.
4491
4492Version 1.6.2rc01 [April 18, 2013]
4493  Added contrib/tools/fixitxt.c, to repair the erroneous iTXt chunk length
4494    written by libpng-1.6.0 and 1.6.1.
4495  Disallow storing sRGB information when the sRGB is not supported.
4496
4497Version 1.6.2rc02 [April 18, 2013]
4498  Merge pngtest.c with libpng-1.7.0
4499
4500Version 1.6.2rc03 [April 22, 2013]
4501  Trivial spelling cleanup.
4502
4503Version 1.6.2rc04 and 1.6.2rc05 [omitted]
4504
4505Version 1.6.2rc06 [April 24, 2013]
4506  Reverted to version 1.6.2rc03.  Recent changes to arm/neon support
4507    have been ported to libpng-1.7.0beta09 and will reappear in version
4508    1.6.3beta01.
4509
4510Version 1.6.2 [April 25, 2013]
4511  No changes.
4512
4513Version 1.6.3beta01 [April 25, 2013]
4514  Revised stack marking in arm/filter_neon.S and configure.ac.
4515  Ensure that NEON filter stuff is completely disabled when switched 'off'.
4516    Previously the ARM NEON specific files were still built if the option
4517    was switched 'off' as opposed to being explicitly disabled.
4518
4519Version 1.6.3beta02 [April 26, 2013]
4520  Test for 'arm*' not just 'arm' in the host_cpu configure variable.
4521  Rebuilt the configure scripts.
4522
4523Version 1.6.3beta03 [April 30, 2013]
4524  Expanded manual paragraph about writing private chunks, particularly
4525    the need to call png_set_keep_unknown_chunks() when writing them.
4526  Avoid dereferencing NULL pointer possibly returned from
4527     png_create_write_struct() (Andrew Church).
4528
4529Version 1.6.3beta05 [May 9, 2013]
4530  Calculate our own zlib windowBits when decoding rather than trusting the
4531    CMF bytes in the PNG datastream.
4532  Added an option to force maximum window size for inflating, which was
4533    the behavior of libpng15 and earlier.
4534  Added png-fix-itxt and png-fix-too-far-back to the built programs and
4535    removed warnings from the source code and timepng that are revealed as
4536    a result.
4537  Detect wrong libpng versions linked to png-fix-too-far-back, which currently
4538    only works with libpng versions that can be made to reliably fail when
4539    the deflate data contains an out-of-window reference.  This means only
4540    1.6 and later.
4541  Fixed gnu issues: g++ needs a static_cast, gcc 4.4.7 has a broken warning
4542    message which it is easier to work round than ignore.
4543  Updated contrib/pngminus/pnm2png.c (Paul Stewart):
4544    Check for EOF
4545    Ignore "#" delimited comments in input file to pnm2png.c.
4546    Fixed whitespace handling
4547    Added a call to png_set_packing()
4548    Initialize dimension values so if sscanf fails at least we have known
4549      invalid values.
4550  Attempt to detect configuration issues with png-fix-too-far-back, which
4551    requires both the correct libpng and the correct zlib to function
4552    correctly.
4553  Check ZLIB_VERNUM for mismatches, enclose #error in quotes
4554  Added information in the documentation about problems with and fixes for
4555    the bad CRC and bad iTXt chunk situations.
4556
4557Version 1.6.3beta06 [May 12, 2013]
4558  Allow contrib/pngminus/pnm2png.c to compile without WRITE_INVERT and
4559    WRITE_PACK supported (writes error message that it can't read P1 or
4560    P4 PBM files).
4561  Improved png-fix-too-far-back usage message, added --suffix option.
4562  Revised contrib/pngminim/*/makefile to generate pnglibconf.h with the
4563    right zlib header files.
4564  Separated CPPFLAGS and CFLAGS in contrib/pngminim/*/makefile
4565
4566Version 1.6.3beta07 [June 8, 2013]
4567  Removed a redundant test in png_set_IHDR().
4568  Added set(CMAKE_CONFIGURATION_TYPES ...) to CMakeLists.txt (Andrew Hundt)
4569  Deleted set(CMAKE_BUILD_TYPE) block from CMakeLists.txt
4570  Enclose the prototypes for the simplified write API in
4571    #ifdef PNG_STDIO_SUPPORTED/#endif
4572  Make ARM NEON support work at compile time (not just configure time).
4573    This moves the test on __ARM_NEON__ into pngconf.h to avoid issues when
4574    using a compiler that compiles for multiple architectures at one time.
4575  Removed PNG_FILTER_OPTIMIZATIONS and PNG_ARM_NEON_SUPPORTED from
4576    pnglibconf.h, allowing more of the decisions to be made internally
4577    (pngpriv.h) during the compile.  Without this, symbol prefixing is broken
4578    under certain circumstances on ARM platforms.  Now only the API parts of
4579    the optimizations ('check' vs 'api') are exposed in the public header files
4580    except that the new setting PNG_ARM_NEON_OPT documents how libpng makes the
4581    decision about whether or not to use the optimizations.
4582  Protect symbol prefixing against CC/CPPFLAGS/CFLAGS useage.
4583    Previous iOS/Xcode fixes for the ARM NEON optimizations moved the test
4584    on __ARM_NEON__ from configure time to compile time.  This breaks symbol
4585    prefixing because the definition of the special png_init_filter_functions
4586    call was hidden at configure time if the relevant compiler arguments are
4587    passed in CFLAGS as opposed to CC.  This change attempts to avoid all
4588    the confusion that would result by declaring the init function even when
4589    it is not used, so that it will always get prefixed.
4590
4591Version 1.6.3beta08 [June 18, 2013]
4592  Revised libpng.3 so that "doclifter" can process it.
4593
4594Version 1.6.3beta09 [June 27, 2013]
4595  Revised example.c to illustrate use of PNG_DEFAULT_sRGB and PNG_GAMMA_MAC_18
4596    as parameters for png_set_gamma().  These have been available since
4597    libpng-1.5.4.
4598  Renamed contrib/tools/png-fix-too-far-back.c to pngfix.c and revised it
4599    to check all compressed chunks known to libpng.
4600
4601Version 1.6.3beta10 [July 5, 2013]
4602  Updated documentation to show default behavior of benign errors correctly.
4603  Only compile ARM code when PNG_READ_SUPPORTED is defined.
4604  Fixed undefined behavior in contrib/tools/pngfix.c and added new strip
4605    option. pngfix relied on undefined behavior and even a simple change from
4606    gcc to g++ caused it to fail.  The new strip option 'unsafe' has been
4607    implemented and is the default if --max is given.  Option names have
4608    been clarified, with --strip=transform now stripping the bKGD chunk,
4609    which was stripped previously with --strip=unused.
4610  Added all documented chunk types to pngpriv.h
4611  Unified pngfix.c source with libpng17.
4612
4613Version 1.6.3rc01 [July 11, 2013]
4614  No changes.
4615
4616Version 1.6.3 [July 18, 2013]
4617  Revised manual about changes in iTXt chunk handling made in libpng-1.6.0.
4618  Added "/* SAFE */" comments in pngrutil.c and pngrtran.c where warnings
4619    may be erroneously issued by code-checking applications.
4620
4621Version 1.6.4beta01 [August 21, 2013]
4622  Added information about png_set_options() to the manual.
4623  Delay calling png_init_filter_functions() until a row with nonzero filter
4624    is found.
4625
4626Version 1.6.4beta02 [August 30, 2013]
4627  Fixed inconsistent conditional compilation of png_chunk_unknown_handling()
4628    prototype, definition, and usage.  Made it depend on
4629    PNG_HANDLE_AS_UNKNOWN_SUPPORTED everywhere.
4630
4631Version 1.6.4rc01 [September 5, 2013]
4632  No changes.
4633
4634Version 1.6.4 [September 12, 2013]
4635  No changes.
4636
4637Version 1.6.5 [September 14, 2013]
4638  Removed two stray lines of code from arm/arm_init.c.
4639
4640Version 1.6.6 [September 16, 2013]
4641  Removed two stray lines of code from arm/arm_init.c, again.
4642
4643Version 1.6.7beta01 [September 30, 2013]
4644  Revised unknown chunk code to correct several bugs in the NO_SAVE_/NO_WRITE
4645    combination
4646  Allow HANDLE_AS_UNKNOWN to work when other options are configured off. Also
4647    fixed the pngminim makefiles to work when $(MAKEFLAGS) contains stuff
4648    which terminates the make options (as by default in recent versions of
4649    Gentoo).
4650  Avoid up-cast warnings in pngvalid.c. On ARM the alignment requirements of
4651    png_modifier are greater than that of png_store and as a consequence
4652    compilation of pngvalid.c results in a warning about increased alignment
4653    requirements because of the bare cast to (png_modifier*). The code is safe,
4654    because the pointer is known to point to a stack allocated png_modifier,
4655    but this change avoids the warning.
4656  Fixed default behavior of ARM_NEON_API. If the ARM NEON API option was
4657    compiled without the CHECK option it defaulted to on, not off.
4658  Check user callback behavior in pngunknown.c. Previous versions compiled
4659    if SAVE_UNKNOWN was not available but did nothing since the callback
4660    was never implemented.
4661  Merged pngunknown.c with 1.7 version and back ported 1.7 improvements/fixes
4662
4663Version 1.6.7beta02 [October 12, 2013]
4664  Made changes for compatibility with automake 1.14:
4665    1) Added the 'compile' program to the list of programs that must be cleaned
4666       in autogen.sh
4667    2) Added 'subdir-objects' which causes .c files in sub-directories to be
4668       compiled such that the corresponding .o files are also in the
4669       sub-directory.  This is because automake 1.14 warns that the
4670       current behavior of compiling to the top level directory may be removed
4671       in the future.
4672    3) Updated dependencies on pnglibconf.h to match the new .o locations and
4673       added all the files in contrib/libtests and contrib/tools that depend
4674       on pnglibconf.h
4675    4) Added 'BUILD_SOURCES = pnglibconf.h'; this is the automake recommended
4676       way of handling the dependencies of sources that are machine generated;
4677       unfortunately it only works if the user does 'make all' or 'make check',
4678       so the dependencies (3) are still required.
4679  Cleaned up (char*) casts of zlib messages. The latest version of the Intel C
4680    compiler complains about casting a string literal as (char*), so copied the
4681    treatment of z_const from the library code into pngfix.c
4682  Simplified error message code in pngunknown. The simplification has the
4683    useful side effect of avoiding a bogus warning generated by the latest
4684    version of the Intel C compiler (it objects to
4685    condition ? string-literal : string-literal).
4686  Make autogen.sh work with automake 1.13 as well as 1.14. Do this by always
4687    removing the 1.14 'compile' script but never checking for it.
4688
4689Version 1.6.7beta03 [October 19, 2013]
4690  Added ARMv8 support (James Yu <james.yu at linaro.org>).  Added file
4691    arm/filter_neon_intrinsics.c; enable with -mfpu=neon.
4692  Revised pngvalid to generate size images with as many filters as it can
4693    manage, limited by the number of rows.
4694  Cleaned up ARM NEON compilation handling. The tests are now in pngpriv.h
4695    and detect the broken GCC compilers.
4696
4697Version 1.6.7beta04 [October 26, 2013]
4698  Allow clang derived from older GCC versions to use ARM intrinsics. This
4699    causes all clang builds that use -mfpu=neon to use the intrinsics code,
4700    not the assembler code.  This has only been tested on iOS 7. It may be
4701    necessary to exclude some earlier clang versions but this seems unlikely.
4702  Changed NEON implementation selection mechanism. This allows assembler
4703    or intrinsics to be turned on at compile time during the build by defining
4704    PNG_ARM_NEON_IMPLEMENTATION to the correct value (2 or 1).  This macro
4705    is undefined by default and the build type is selected in pngpriv.h.
4706
4707Version 1.6.7rc01 [November 2, 2013]
4708  No changes.
4709
4710Version 1.6.7rc02 [November 7, 2013]
4711  Fixed #include in filter_neon_intrinsics.c and ctype macros. The ctype char
4712    checking macros take an unsigned char argument, not a signed char.
4713
4714Version 1.6.7 [November 14, 2013]
4715  No changes.
4716
4717Version 1.6.8beta01 [November 24, 2013]
4718  Moved prototype for png_handle_unknown() in pngpriv.h outside of
4719    the #ifdef PNG_SET_UNKNOWN_CHUNKS_SUPPORTED/#endif block.
4720  Added "-Wall" to CFLAGS in contrib/pngminim/*/makefile
4721  Conditionally compile some unused functions reported by -Wall in
4722    pngminim.
4723  Fixed 'minimal' builds. Various obviously useful minimal configurations
4724    don't build because of missing contrib/libtests test programs and
4725    overly complex dependencies in scripts/pnglibconf.dfa. This change
4726    adds contrib/conftest/*.dfa files that can be used in automatic build
4727    scripts to ensure that these configurations continue to build.
4728  Enabled WRITE_INVERT and WRITE_PACK in contrib/pngminim/encoder.
4729  Fixed pngvalid 'fail' function declaration on the Intel C Compiler.
4730    This reverts to the previous 'static' implementation and works round
4731    the 'unused static function' warning by using PNG_UNUSED().
4732
4733Version 1.6.8beta02 [November 30, 2013]
4734  Removed or marked PNG_UNUSED some harmless "dead assignments" reported
4735    by clang scan-build.
4736  Changed tabs to 3 spaces in png_debug macros and changed '"%s"m'
4737    to '"%s" m' to improve portability among compilers.
4738  Changed png_free_default() to free() in pngtest.c
4739
4740Version 1.6.8rc01 [December 12, 2013]
4741  Tidied up pngfix inits and fixed pngtest no-write builds.
4742
4743Version 1.6.8rc02 [December 14, 2013]
4744  Handle zero-length PLTE chunk or NULL palette with png_error()
4745    instead of png_chunk_report(), which by default issues a warning
4746    rather than an error, leading to later reading from a NULL pointer
4747    (png_ptr->palette) in png_do_expand_palette(). This is CVE-2013-6954
4748    and VU#650142.  Libpng-1.6.1 through 1.6.7 are vulnerable.
4749    Libpng-1.6.0 and earlier do not have this bug.
4750
4751Version 1.6.8 [December 19, 2013]
4752  No changes.
4753
4754Version 1.6.9beta01 [December 26, 2013]
4755  Bookkeeping: Moved functions around (no changes). Moved transform
4756    function definitions before the place where they are called so that
4757    they can be made static. Move the intrapixel functions and the
4758    grayscale palette builder out of the png?tran.c files. The latter
4759    isn't a transform function and is no longer used internally, and the
4760    former MNG specific functions are better placed in pngread/pngwrite.c
4761  Made transform implementation functions static. This makes the internal
4762    functions called by png_do_{read|write}_transformations static. On an
4763    x86-64 DLL build (Gentoo Linux) this reduces the size of the text
4764    segment of the DLL by 1208 bytes, about 0.6%. It also simplifies
4765    maintenance by removing the declarations from pngpriv.h and allowing
4766    easier changes to the internal interfaces.
4767  Rebuilt configure scripts with automake-1.14.1 and autoconf-2.69
4768    in the tar distributions.
4769
4770Version 1.6.9beta02 [January 1, 2014]
4771  Added checks for libpng 1.5 to pngvalid.c.  This supports the use of
4772    this version of pngvalid in libpng 1.5
4773  Merged with pngvalid.c from libpng-1.7 changes to create a single
4774    pngvalid.c
4775  Removed #error macro from contrib/tools/pngfix.c (Thomas Klausner).
4776  Merged pngrio.c, pngtrans.c, pngwio.c, and pngerror.c with libpng-1.7.0
4777  Merged libpng-1.7.0 changes to make no-interlace configurations work
4778    with test programs.
4779  Revised pngvalid.c to support libpng 1.5, which does not support the
4780    PNG_MAXIMUM_INFLATE_WINDOW option, so #define it out when appropriate in
4781    pngvalid.c
4782  Allow unversioned links created on install to be disabled in configure.
4783    In configure builds 'make install' changes/adds links like png.h
4784    and libpng.a to point to the newly installed, versioned, files (e.g.
4785    libpng17/png.h and libpng17.a). Three new configure options and some
4786    rearrangement of Makefile.am allow creation of these links to be disabled.
4787
4788Version 1.6.9beta03 [January 10, 2014]
4789  Removed potentially misleading warning from png_check_IHDR().
4790
4791Version 1.6.9beta04 [January 20, 2014]
4792  Updated scripts/makefile.* to use CPPFLAGS (Cosmin).
4793  Added clang attribute support (Cosmin).
4794
4795Version 1.6.9rc01 [January 28, 2014]
4796  No changes.
4797
4798Version 1.6.9rc02 [January 30, 2014]
4799  Quiet an uninitialized memory warning from VC2013 in png_get_png().
4800
4801Version 1.6.9 [February 6, 2014]
4802
4803Version 1.6.10beta01 [February 9, 2014]
4804  Backported changes from libpng-1.7.0beta30 and beta31:
4805  Fixed a large number of instances where PNGCBAPI was omitted from
4806    function definitions.
4807  Added pngimage test program for png_read_png() and png_write_png()
4808    with two new test scripts.
4809  Removed dependence on !PNG_READ_EXPAND_SUPPORTED for calling
4810    png_set_packing() in png_read_png().
4811  Fixed combination of ~alpha with shift. On read invert alpha, processing
4812    occurred after shift processing, which causes the final values to be
4813    outside the range that should be produced by the shift. Reversing the
4814    order on read makes the two transforms work together correctly and mirrors
4815    the order used on write.
4816  Do not read invalid sBIT chunks. Previously libpng only checked sBIT
4817    values on write, so a malicious PNG writer could therefore cause
4818    the read code to return an invalid sBIT chunk, which might lead to
4819    application errors or crashes.  Such chunks are now skipped (with
4820    chunk_benign_error).
4821  Make png_read_png() and png_write_png() prototypes in png.h depend
4822    upon PNG_READ_SUPPORTED and PNG_WRITE_SUPPORTED.
4823  Support builds with unsupported PNG_TRANSFORM_* values.  All of the
4824    PNG_TRANSFORM_* values are always defined in png.h and, because they
4825    are used for both read and write in some cases, it is not reliable
4826    to #if out ones that are totally unsupported. This change adds error
4827    detection in png_read_image() and png_write_image() to do a
4828    png_app_error() if the app requests something that cannot be done
4829    and it adds corresponding code to pngimage.c to handle such options
4830    by not attempting to test them.
4831
4832Version 1.6.10beta02 [February 23, 2014]
4833  Moved redefines of png_error(), png_warning(), png_chunk_error(),
4834    and png_chunk_warning() from pngpriv.h to png.h to make them visible
4835    to libpng-calling applications.
4836  Moved OS dependent code from arm/arm_init.c, to allow the included
4837    implementation of the ARM NEON discovery function to be set at
4838    build-time and provide sample implementations from the current code in the
4839    contrib/arm-neon subdirectory. The __linux__ code has also been changed to
4840    compile and link on Android by using /proc/cpuinfo, and the old linux code
4841    is in contrib/arm-neon/linux-auxv.c.  The new code avoids POSIX and Linux
4842    dependencies apart from opening /proc/cpuinfo and is C90 compliant.
4843  Check for info_ptr == NULL early in png_read_end() so we don't need to
4844    run all the png_handle_*() and depend on them to return if info_ptr == NULL.
4845    This improves the performance of png_read_end(png_ptr, NULL) and makes
4846    it more robust against future programming errors.
4847  Check for __has_extension before using it in pngconf.h, to
4848    support older Clang versions (Jeremy Sequoia).
4849  Treat CRC error handling with png_set_crc_action(), instead of with
4850    png_set_benign_errors(), which has been the case since libpng-1.6.0beta18.
4851  Use a user warning handler in contrib/gregbook/readpng2.c instead of default,
4852    so warnings will be put on stderr even if libpng has CONSOLE_IO disabled.
4853  Added png_ptr->process_mode = PNG_READ_IDAT_MODE in png_push_read_chunk
4854    after recognizing the IDAT chunk, which avoids an infinite loop while
4855    reading a datastream whose first IDAT chunk is of zero-length.
4856    This fixes CERT VU#684412 and CVE-2014-0333.
4857  Don't recognize known sRGB profiles as sRGB if they have been hacked,
4858    but don't reject them and don't issue a copyright violation warning.
4859
4860Version 1.6.10beta03 [February 25, 2014]
4861  Moved some documentation from png.h to libpng.3 and libpng-manual.txt
4862  Minor editing of contrib/arm-neon/README and contrib/examples/*.c
4863
4864Version 1.6.10rc01 [February 27, 2014]
4865  Fixed typos in the manual and in scripts/pnglibconf.dfa (CFLAGS -> CPPFLAGS
4866    and PNG_USR_CONFIG -> PNG_USER_CONFIG).
4867
4868Version 1.6.10rc02 [February 28, 2014]
4869  Removed unreachable return statement after png_chunk_error()
4870    in pngrutil.c
4871
4872Version 1.6.10rc03 [March 4, 2014]
4873  Un-deprecated png_data_freer().
4874
4875Version 1.6.10 [March 6, 2014]
4876
4877Send comments/corrections/commendations to png-mng-implement at lists.sf.net
4878(subscription required; visit
4879https://lists.sourceforge.net/lists/listinfo/png-mng-implement
4880to subscribe)
4881or to glennrp at users.sourceforge.net
4882
4883Glenn R-P
4884#endif
4885