7 * liblzma: Fix lzma_index_init(). It could crash if memory allocation
10 * liblzma: Fix the possibility of an incorrect LZMA_BUF_ERROR when
11 a BCJ filter is used and the application only provides exactly
12 as much output space as is the uncompressed size of the file.
14 * Fix "xz -lvv foo.xz". It could crash on some corrupted files.
16 * Fix output of "xz --robot -lv" and "xz --robot -lvv" which
17 incorrectly printed the filename also in the "foo (x/x)" format.
19 * Fix exit status of "xzdiff foo.xz bar.xz".
21 * Fix exit status of "xzgrep foo binary_file".
23 * Fix portability to EBCDIC systems.
25 * Fix a configure issue on AIX with the XL C compiler. See INSTALL
33 - A memory leak was fixed.
35 - lzma_stream_buffer_encode() no longer creates an empty .xz
36 Block if encoding an empty buffer. Such an empty Block with
37 LZMA2 data would trigger a bug in 5.0.1 and older (see the
38 first bullet point in 5.0.2 notes). When releasing 5.0.2,
39 I thought that no encoder creates this kind of files but
42 - Validate function arguments better in a few functions. Most
43 importantly, specifying an unsupported integrity check to
44 lzma_stream_buffer_encode() no longer creates a corrupt .xz
45 file. Probably no application tries to do that, so this
46 shouldn't be a big problem in practice.
48 - Document that lzma_block_buffer_encode(),
49 lzma_easy_buffer_encode(), lzma_stream_encoder(), and
50 lzma_stream_buffer_encode() may return LZMA_UNSUPPORTED_CHECK.
52 - The return values of the _memusage() functions are now
55 * Fix command name detection in xzgrep. xzegrep and xzfgrep now
56 correctly use egrep and fgrep instead of grep.
58 * French translation was added.
63 * LZMA2 decompressor now correctly accepts LZMA2 streams with no
64 uncompressed data. Previously it considered them corrupt. The
65 bug can affect applications that use raw LZMA2 streams. It is
66 very unlikely to affect .xz files because no compressor creates
67 .xz files with empty LZMA2 streams. (Empty .xz files are a
68 different thing than empty LZMA2 streams.)
70 * "xz --suffix=.foo filename.foo" now refuses to compress the
71 file due to it already having the suffix .foo. It was already
72 documented on the man page, but the code lacked the test.
74 * "xzgrep -l foo bar.xz" works now.
76 * Polish translation was added.
81 * xz --force now (de)compresses files that have setuid, setgid,
82 or sticky bit set and files that have multiple hard links.
83 The man page had it documented this way already, but the code
86 * gzip and bzip2 support in xzdiff was fixed.
90 * Minor fix to Czech translation
95 Only the most important changes compared to 4.999.9beta are listed
96 here. One change is especially important:
98 * The memory usage limit is now disabled by default. Some scripts
99 written before this change may have used --memory=max on xz command
100 line or in XZ_OPT. THESE USES OF --memory=max SHOULD BE REMOVED
101 NOW, because they interfere with user's ability to set the memory
102 usage limit himself. If user-specified limit causes problems to
103 your script, blame the user.
105 Other significant changes:
107 * Added support for XZ_DEFAULTS environment variable. This variable
108 allows users to set default options for xz, e.g. default memory
109 usage limit or default compression level. Scripts that use xz
110 must never set or unset XZ_DEFAULTS. Scripts should use XZ_OPT
111 instead if they need a way to pass options to xz via an
112 environment variable.
114 * The compression settings associated with the preset levels
115 -0 ... -9 have been changed. --extreme was changed a little too.
116 It is now less likely to make compression worse, but with some
117 files the new --extreme may compress slightly worse than the old
120 * If a preset level (-0 ... -9) is specified after a custom filter
121 chain options have been used (e.g. --lzma2), the custom filter
122 chain will be forgotten. Earlier the preset options were
123 completely ignored after custom filter chain options had been
126 * xz will create sparse files when decompressing if the uncompressed
127 data contains long sequences of binary zeros. This is done even
128 when writing to standard output that is connected to a regular
129 file and certain additional conditions are met to make it safe.
131 * Support for "xz --list" was added. Combine with --verbose or
132 --verbose --verbose (-vv) for detailed output.
134 * I had hoped that liblzma API would have been stable after
135 4.999.9beta, but there have been a couple of changes in the
136 advanced features, which don't affect most applications:
138 - Index handling code was revised. If you were using the old
139 API, you will get a compiler error (so it's easy to notice).
141 - A subtle but important change was made to the Block handling
142 API. lzma_block.version has to be initialized even for
143 lzma_block_header_decode(). Code that doesn't do it will work
144 for now, but might break in the future, which makes this API
147 * The major soname has been bumped to 5.0.0. liblzma API and ABI
148 are now stable, so the need to recompile programs linking against
149 liblzma shouldn't arise soon.