diff options
Diffstat (limited to 'ChangeLog')
| -rw-r--r-- | ChangeLog | 68 |
1 files changed, 59 insertions, 9 deletions
| @@ -1,20 +1,54 @@ | |||
| 1 | 1 | ||
| 2 | ChangeLog file for zlib | 2 | ChangeLog file for zlib |
| 3 | 3 | ||
| 4 | Changes in 1.0 (15 Jan 96) | 4 | Changes in 1.0.1 (20 May 96) [1.0 skipped to avoid confusion] |
| 5 | - fix array overlay in deflate.c which sometimes caused bad compressed data | ||
| 6 | - fix inflate bug with empty stored block | ||
| 7 | - fix MSDOS medium model which was broken in 0.99 | ||
| 8 | - fix deflateParams() which could generated bad compressed data. | ||
| 9 | - Bytef is define'd instead of typedef'ed (work around Borland bug) | ||
| 10 | - added an INDEX file | ||
| 11 | - new makefiles for DJGPP (Makefile.dj2), 32-bit Borland (Makefile.b32), | ||
| 12 | Watcom (Makefile.wat), Amiga SAS/C (Makefile.sas) | ||
| 13 | - speed up adler32 for modern machines without auto-increment | ||
| 14 | - added -ansi for IRIX in configure | ||
| 15 | - static_init_done in trees.c is an int | ||
| 16 | - define unlink as delete for VMS | ||
| 17 | - fix configure for QNX | ||
| 18 | - add configure branch for SCO and HPUX | ||
| 19 | - avoid many warnings (unused variables, dead assignments, etc...) | ||
| 20 | - no fdopen for BeOS | ||
| 21 | - fix the Watcom fix for 32 bit mode (define FAR as empty) | ||
| 22 | - removed redefinition of Byte for MKWERKS | ||
| 23 | - work around an MWKERKS bug (incorrect merge of all .h files) | ||
| 24 | |||
| 25 | Changes in 0.99 (27 Jan 96) | ||
| 26 | - allow preset dictionary shared between compressor and decompressor | ||
| 27 | - allow compression level 0 (no compression) | ||
| 28 | - add deflateParams in zlib.h: allow dynamic change of compression level | ||
| 29 | and compression strategy. | ||
| 30 | - test large buffers and deflateParams in example.c | ||
| 31 | - add optional "configure" to build zlib as a shared library | ||
| 32 | - suppress Makefile.qnx, use configure instead | ||
| 5 | - fixed deflate for 64-bit systems (detected on Cray) | 33 | - fixed deflate for 64-bit systems (detected on Cray) |
| 6 | - fixed inflate_blocks for 64-bit systems (detected on Alpha) | 34 | - fixed inflate_blocks for 64-bit systems (detected on Alpha) |
| 35 | - declare Z_DEFLATED in zlib.h (possible parameter for deflateInit2) | ||
| 36 | - always return Z_BUF_ERROR when deflate() has nothing to do | ||
| 37 | - deflateInit and inflateInit are now macros to allow version checking | ||
| 38 | - prefix all global functions and types with z_ with -DZ_PREFIX | ||
| 7 | - make falloc completely reentrant (inftrees.c) | 39 | - make falloc completely reentrant (inftrees.c) |
| 8 | - fixed very unlikely race condition in ct_static_init | 40 | - fixed very unlikely race condition in ct_static_init |
| 9 | - always return Z_BUF_ERROR when deflate() has nothing to do | ||
| 10 | - add deflateParams in zlib.h. | ||
| 11 | - test large buffers and deflateParams in example.c | ||
| 12 | - declare Z_DEFLATED in zlib.h (possible parameter for deflateInit2) | ||
| 13 | - Document Z_SYNC_FLUSH | ||
| 14 | - added deflateGetAdler32 and deflateSetAdler32 | ||
| 15 | - free in reverse order of allocation to help memory manager | 41 | - free in reverse order of allocation to help memory manager |
| 42 | - use zlib-1.0/* instead of zlib/* inside the tar.gz | ||
| 43 | - make zlib warning-free with "gcc -O3 -Wall -Wwrite-strings -Wpointer-arith | ||
| 44 | -Wconversion -Wstrict-prototypes -Wmissing-prototypes" | ||
| 45 | - allow gzread on concatenated .gz files | ||
| 46 | - deflateEnd now returns Z_DATA_ERROR if it was premature | ||
| 47 | - deflate is finally (?) fully deterministic (no matches beyond end of input) | ||
| 48 | - Document Z_SYNC_FLUSH | ||
| 49 | - add uninstall in Makefile | ||
| 16 | - Check for __cpluplus in zlib.h | 50 | - Check for __cpluplus in zlib.h |
| 17 | - Finer test in ct_align for partial flush | 51 | - Better test in ct_align for partial flush |
| 18 | - avoid harmless warnings for Borland C++ | 52 | - avoid harmless warnings for Borland C++ |
| 19 | - initialize hash_head in deflate.c | 53 | - initialize hash_head in deflate.c |
| 20 | - avoid warning on fdopen (gzio.c) for HP cc -Aa | 54 | - avoid warning on fdopen (gzio.c) for HP cc -Aa |
| @@ -23,7 +57,7 @@ Changes in 1.0 (15 Jan 96) | |||
| 23 | - ignore error if ranlib doesn't exist | 57 | - ignore error if ranlib doesn't exist |
| 24 | - call ranlib twice for NeXTSTEP | 58 | - call ranlib twice for NeXTSTEP |
| 25 | - use exec_prefix instead of prefix for libz.a | 59 | - use exec_prefix instead of prefix for libz.a |
| 26 | - renamed ct_* as tr_* to avoid conflict with another application. | 60 | - renamed ct_* as _tr_* to avoid conflict with applications |
| 27 | - clear z->msg in inflateInit2 before any error return | 61 | - clear z->msg in inflateInit2 before any error return |
| 28 | - initialize opaque in example.c, gzio.c, deflate.c and inflate.c | 62 | - initialize opaque in example.c, gzio.c, deflate.c and inflate.c |
| 29 | - fixed typo in zconf.h (_GNUC__ => __GNUC__) | 63 | - fixed typo in zconf.h (_GNUC__ => __GNUC__) |
| @@ -32,6 +66,22 @@ Changes in 1.0 (15 Jan 96) | |||
| 32 | - in fcalloc, normalize pointer if size > 65520 bytes | 66 | - in fcalloc, normalize pointer if size > 65520 bytes |
| 33 | - don't use special fcalloc for 32 bit Borland C++ | 67 | - don't use special fcalloc for 32 bit Borland C++ |
| 34 | - use STDC instead of __GO32__ to avoid redeclaring exit, calloc, etc... | 68 | - use STDC instead of __GO32__ to avoid redeclaring exit, calloc, etc... |
| 69 | - use Z_BINARY instead of BINARY | ||
| 70 | - document that gzclose after gzdopen will close the file | ||
| 71 | - allow "a" as mode in gzopen. | ||
| 72 | - fix error checking in gzread | ||
| 73 | - allow skipping .gz extra-field on pipes | ||
| 74 | - added reference to Perl interface in README | ||
| 75 | - put the crc table in FAR data (I dislike more and more the medium model :) | ||
| 76 | - added get_crc_table | ||
| 77 | - added a dimension to all arrays (Borland C can't count). | ||
| 78 | - workaround Borland C bug in declaration of inflate_codes_new & inflate_fast | ||
| 79 | - guard against multiple inclusion of *.h (for precompiled header on Mac) | ||
| 80 | - Watcom C pretends to be Microsoft C small model even in 32 bit mode. | ||
| 81 | - don't use unsized arrays to avoid silly warnings by Visual C++: | ||
| 82 | warning C4746: 'inflate_mask' : unsized array treated as '__far' | ||
| 83 | (what's wrong with far data in far model?). | ||
| 84 | - define enum out of inflate_blocks_state to allow compilation with C++ | ||
| 35 | 85 | ||
| 36 | Changes in 0.95 (16 Aug 95) | 86 | Changes in 0.95 (16 Aug 95) |
| 37 | - fix MSDOS small and medium model (now easier to adapt to any compiler) | 87 | - fix MSDOS small and medium model (now easier to adapt to any compiler) |
