ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/liblzf/Changes
Revision: 1.47
Committed: Mon Dec 14 02:05:14 2009 UTC (14 years, 5 months ago) by root
Branch: MAIN
Changes since 1.46: +3 -0 lines
Log Message:
*** empty log message ***

File Contents

# Content
1
2 - fixed hash calculation in C♯ version (Tiago Freitas Leal).
3
4 3.5 Fri May 1 02:28:42 CEST 2009
5 - lzf_compress did sometimes write one octet past the given output
6 buffer (analyzed and nice testcase by Salvatore Sanfilippo).
7
8 3.4 Tue Sep 2 06:45:00 CEST 2008
9 - the fix from 3.3 introduced a compression bug, which is fixed in
10 this release (which explains the mysterious prerelease...). Thanks
11 once more to Clément Calmels.
12
13 3.3 Mon Aug 25 03:17:42 CEST 2008
14 - lzf_compress could access memory after the given input buffer
15 when outputting back references. reported with nice testcase
16 by Clément Calmels.
17
18 3.2 Fri May 9 18:52:23 CEST 2008
19 - include a workaround for failing POSIX and real-world compliance
20 on 64 bit windows (microsoft claims to support POSIX, but is far
21 from it). (bug found and analysed nicely by John Lilley).
22
23 3.1 Fri Nov 30 11:33:04 CET 2007
24 - IMPORTANT BUGFIX: a too long final literal run would corrupt data
25 in the encoder (this was introduced in 3.0 only, earlier versions
26 are safe).
27
28 3.0 Tue Nov 13 22:13:09 CET 2007
29 - switched to 2-clause bsd with "GPL v2 or any later version" option.
30 - speed up compression by ~10-15% in common cases
31 by some manual unrolling.
32 - import some compiler tricks from JSON::XS, for further speed-ups.
33 - tune hash functions depending on ULTRA_FAST or VERY_FAST settings.
34 - for typical binary data (e.g. /bin/bash, memory dumps,
35 canterbury corpus etc.), speed is now comparable to fastlz, but
36 with better compression ratio. with ULTRA_FAST, it's typically
37 3-15% faster than fastlz while still maintaining a similar ratio.
38 (amd64 and core 2 duo, ymmv). thanks a lot for the competition :)
39 - undo inline assembly in compressor, it is no longer helpful.
40 - no changes to the decompressor.
41 - use a HLOG of 16 by default now (formerly 15).
42
43 2.1 Fri Nov 2 13:34:42 CET 2007
44 - switched to a 2-clause bsd license with GPL exception.
45 - get rid of memcpy.
46 - tentatively use rep movsb on x86 and x86_64 (gcc only) for a
47 moderate speed improvement.
48 - applied patch by Kein-Hong Man to maske lzf.c compile under
49 the crippled mingw32 environment.
50
51 2.0 Fri Feb 16 23:11:18 CET 2007
52 - replaced lzf demo by industrial-strength lzf utility with behaviour
53 similar other compression utilities. Thanks for Stefan Traby for
54 rewriting it!
55 - fix state arg prototype.
56
57 1.7 Wed Sep 27 17:29:15 CEST 2006
58 - remove bogus "unlzf" patch.
59 note to self: never accept well-meant patches.
60 - make lzf more robust in presence of padding bytes or sudden eof.
61
62 1.6 Fri Jul 7 17:31:26 CEST 2006
63 - the lzf example utility will now uncompress if invoked
64 as "unlzf" (patch by Scott Feeney).
65 - add CHECK_INPUT option that adds more checks for input
66 data validity.
67 - help applications that do not pass in the correct length
68 (such as php) by returning either EINVAL or E2BIG.
69 - default HLOG size is now 15 (cpu caches have increased).
70 - documentation fixes.
71
72 1.51 Thu Apr 14 22:15:46 CEST 2005
73 - incorporated C♯ implementation of both the en- and decoder,
74 written by "Oren J. Maurice".
75 You can find it in the cs/ subdirectory.
76 - make FRST, NEXT IDX overridable if lzf_c.c is directly included
77 in the code.
78
79 1.5 Tue Mar 8 20:23:23 CET 2005
80 - incorporated improvements by Adam D. Moss,
81 which includes a new VERY_FAST mode which is
82 a bit slower than ULTRA_FAST but much better,
83 and enabled it as default.
84
85 1.401 Thu Mar 3 18:00:52 CET 2005
86 - use cstring in c++, not string.h.
87 - change of contact address.
88
89 1.4 Wed Dec 15 08:08:49 CET 2004
90 - very very slight tuning of the hashing function.
91
92 1.3 Thu Mar 25 15:41:17 CET 2004
93 - changed license of lzf core code to explicitly allow
94 relicensing under the GPLv2.
95 - added VPATH support as suggested by Björn Eriksson.
96
97 1.2 Mon Dec 29 13:47:28 CET 2003
98 - avoid spurious memory accesses after the to-be-compressed
99 memory region. originally reported by Michal Zalewski.
100 - flip LZF_STACK_ARG meaning (to be correct).
101
102 1.1 Tue Dec 23 05:48:32 CET 2003
103 - removed #warn directive, it's not worth the hassle.
104 - add LZF_STACK_ARG and AVOID_ERRNO configurations
105 for embedded systems.
106 - make it compile cleanly as c++.
107 - some small documentation and code fixes.
108
109 1.0 Sun Nov 17 12:37:37 CET 2002
110 - slightly better compression ratio, almost unmeasurably
111 slower.
112 - some documentation fixes.
113
114 0.4 Thu Jun 13 14:11:10 CEST 2002
115 - typoe fix.
116 - lzf demo program now properly decompresses small files.
117 - fix another 64 bit issue, found by Laurent Deniel.
118
119 0.3 Tue Jan 16 13:21:14 CET 2001
120 - fix silly beginners 32/64 bit mistake.
121
122 0.2 Thu Jan 4 05:56:42 CET 2001
123 - now totally independent of autoconfig, for
124 easy inclusion into other programs.
125 - much better fine-tuning, faster and better than 0.1.
126
127 0.1 2000
128 - initial release.