1 2XZ Utils To-Do List 3=================== 4 5Known bugs 6---------- 7 8 The test suite is too incomplete. 9 10 If the memory usage limit is less than about 13 MiB, xz is unable to 11 automatically scale down the compression settings enough even though 12 it would be possible by switching from BT2/BT3/BT4 match finder to 13 HC3/HC4. 14 15 XZ Utils compress some files significantly worse than LZMA Utils. 16 This is due to faster compression presets used by XZ Utils, and 17 can often be worked around by using "xz --extreme". With some files 18 --extreme isn't enough though: it's most likely with files that 19 compress extremely well, so going from compression ratio of 0.003 20 to 0.004 means big relative increase in the compressed file size. 21 22 xz doesn't quote unprintable characters when it displays file names 23 given on the command line. 24 25 tuklib_exit() doesn't block signals => EINTR is possible. 26 27 If liblzma has created threads and fork() gets called, liblzma 28 code will break in the child process unless it calls exec() and 29 doesn't touch liblzma. 30 31 32Missing features 33---------------- 34 35 Add support for storing metadata in .xz files. A preliminary 36 idea is to create a new Stream type for metadata. When both 37 metadata and data are wanted in the same .xz file, two or more 38 Streams would be concatenated. 39 40 The state stored in lzma_stream should be cloneable, which would 41 be mostly useful when using a preset dictionary in LZMA2, but 42 it may have other uses too. Compare to deflateCopy() in zlib. 43 44 Support LZMA_FINISH in raw decoder to indicate end of LZMA1 and 45 other streams that don't have an end of payload marker. 46 47 Adjust dictionary size when the input file size is known. 48 Maybe do this only if an option is given. 49 50 xz doesn't support copying extended attributes, access control 51 lists etc. from source to target file. 52 53 Multithreaded compression: 54 - Reduce memory usage of the current method. 55 - Implement threaded match finders. 56 - Implement pigz-style threading in LZMA2. 57 58 Buffer-to-buffer coding could use less RAM (especially when 59 decompressing LZMA1 or LZMA2). 60 61 I/O library is not implemented (similar to gzopen() in zlib). 62 It will be a separate library that supports uncompressed, .gz, 63 .bz2, .lzma, and .xz files. 64 65 Support changing lzma_options_lzma.mode with lzma_filters_update(). 66 67 Support LZMA_FULL_FLUSH for lzma_stream_decoder() to stop at 68 Block and Stream boundaries. 69 70 lzma_strerror() to convert lzma_ret to human readable form? 71 This is tricky, because the same error codes are used with 72 slightly different meanings, and this cannot be fixed anymore. 73 74 Make it possible to adjust LZMA2 options in the middle of a Block 75 so that the encoding speed vs. compression ratio can be optimized 76 when the compressed data is streamed over network. 77 78 Improved BCJ filters. The current filters are small but they aren't 79 so great when compressing binary packages that contain various file 80 types. Specifically, they make things worse if there are static 81 libraries or Linux kernel modules. The filtering could also be 82 more effective (without getting overly complex), for example, 83 streamable variant BCJ2 from 7-Zip could be implemented. 84 85 Filter that autodetects specific data types in the input stream 86 and applies appropriate filters for the corrects parts of the input. 87 Perhaps combine this with the BCJ filter improvement point above. 88 89 Long-range LZ77 method as a separate filter or as a new LZMA2 90 match finder. 91 92 93Documentation 94------------- 95 96 More tutorial programs are needed for liblzma. 97 98 Document the LZMA1 and LZMA2 algorithms. 99 100 101Miscellaneous 102------------ 103 104 Try to get the media type for .xz registered at IANA. 105 106