Lzip is a lossless data compressor based on the LZMA algorithm, with very safe integrity checking and a user interface similar to the one of gzip or bzip2. Lzip decompresses almost as quickly as gzip and compresses better than bzip2, which makes it well suited for software distribution and data archiving. If you ever need to recover data from a damaged lzip file, try the lziprecover program.
| Tags | Archiving Compression |
|---|---|
| Licenses | GPLv3+ |
| Operating Systems | POSIX |
| Implementation | C++ |
Recent releases


Release Notes: Lziprecover has been moved to its own package. The inability to change output file attributes has been downgraded from an error to a warning. The compression time of the "-0" option has been reduced by 2%. The compression code was reorganized. A small change has been made in the "--help" output and man page. Quote characters in messages have been changed as advised by GNU coding standards. The "--datadir" configure option has been renamed to "--datarootdir" to follow GNU standards.


Release Notes: This release adds minor fixes and cleanups.


Release Notes: The compression time of the option "-0" has been reduced by 2%. A reorganization of the compression code has been made. A small change has been made in the "--help" output and the man page. Quote characters in messages have been changed as advised by GNU Coding Standards.


Release Notes: Lziprecover has been moved to its own package. The inability to change output file attributes has been downgraded from error to warning.


Release Notes: The option "-F, --recompress", which forces recompression of files whose name already has the ".lz" or ".tlz" suffix, has been added. For large values of "--match-length", the compression ratio has been slightly increased, and compression time has been reduced by up to 6%. The compression time of the option "-0" has been reduced by 2%. This release will print only one status line for each multi-member file when only one "-v" is specified. It will print up to 6 bytes of trailing garbage when "-vvvv" is specified. It does not show the message "and `--stdout' was not specified" for file types that can't be read (directories, etc).
Recent comments
09 Dec 2008 11:56
Re: Other LZMA tools
I think you are right that the standalone 'lzma' program (replacing the older lzmash) has a very basic data format. But still, it works, and is the more established tool. I would be happy for lzip to replace it if lzip is better, but to do that it should include support for decompressing legacy .lzma files.
(I note that the gzip format has provision for alternative compression methods but nobody ever seems to use it.)
> As for lrzip, it is actually
> an extension of rzip---and the two are
> more of a proof-of-concept than a
> realworld-workable format.
The file format may be basic but the tool is very good. It usually compresses better than plain LZMA (the algorithm, used in both lzma-utils and lzip) and faster too. LZMA is better for all-purpose use but for batch compression tasks where you don't mind relatively high memory usage, lrzip can give a big improvement. For some Subversion dump files I back up overnight it gave a fourfold increase in compression for about the same speed.
24 Nov 2008 21:05
Re: Other LZMA tools
As I gather, lzma-utils-produced files lack magic identification bytes and a checksum, and if you believe forum archives, lzma-utils did not manage to come up with a suitable new format in more than two years. It is about time lzip came along—7z sounds nice too, but seems to have gotten no ground in the Unix world due to subpar unix integration.
As for lrzip, it is actually an extension of rzip—and the two are more of a proof-of-concept than a realworld-workable format.
01 Sep 2008 16:35
Other LZMA tools
An alternative command-line compressor using LZMA is lzmash from lzma-utils. Also interesting is lrzip which adds a preprocessing step before the LZMA compression to give a better speed/tightness tradeoff, at the expense of using more memory.