Home | History | Annotate | Line # | Download | only in dist 
      1 
      2 XZ Utils To-Do List
      3 ===================
      4 
      5 Known 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     SIGTSTP is not handled. If xz is stopped, the estimated remaining
     28     time and calculated (de)compression speed won't make sense in the
     29     progress indicator (xz --verbose).
     30 
     31     If liblzma has created threads and fork() gets called, liblzma
     32     code will break in the child process unless it calls exec() and
     33     doesn't touch liblzma.
     34 
     35 
     36 Missing features
     37 ----------------
     38 
     39     Add support for storing metadata in .xz files. A preliminary
     40     idea is to create a new Stream type for metadata. When both
     41     metadata and data are wanted in the same .xz file, two or more
     42     Streams would be concatenated.
     43 
     44     The state stored in lzma_stream should be cloneable, which would
     45     be mostly useful when using a preset dictionary in LZMA2, but
     46     it may have other uses too. Compare to deflateCopy() in zlib.
     47 
     48     Support LZMA_FINISH in raw decoder to indicate end of LZMA1 and
     49     other streams that don't have an end of payload marker.
     50 
     51     Adjust dictionary size when the input file size is known.
     52     Maybe do this only if an option is given.
     53 
     54     xz doesn't support copying extended attributes, access control
     55     lists etc. from source to target file.
     56 
     57     Multithreaded compression:
     58       - Reduce memory usage of the current method.
     59       - Implement threaded match finders.
     60       - Implement pigz-style threading in LZMA2.
     61 
     62     Multithreaded decompression
     63 
     64     Buffer-to-buffer coding could use less RAM (especially when
     65     decompressing LZMA1 or LZMA2).
     66 
     67     I/O library is not implemented (similar to gzopen() in zlib).
     68     It will be a separate library that supports uncompressed, .gz,
     69     .bz2, .lzma, and .xz files.
     70 
     71     Support changing lzma_options_lzma.mode with lzma_filters_update().
     72 
     73     Support LZMA_FULL_FLUSH for lzma_stream_decoder() to stop at
     74     Block and Stream boundaries.
     75 
     76     lzma_strerror() to convert lzma_ret to human readable form?
     77     This is tricky, because the same error codes are used with
     78     slightly different meanings, and this cannot be fixed anymore.
     79 
     80     Make it possible to adjust LZMA2 options in the middle of a Block
     81     so that the encoding speed vs. compression ratio can be optimized
     82     when the compressed data is streamed over network.
     83 
     84     Improved BCJ filters. The current filters are small but they aren't
     85     so great when compressing binary packages that contain various file
     86     types. Specifically, they make things worse if there are static
     87     libraries or Linux kernel modules. The filtering could also be
     88     more effective (without getting overly complex), for example,
     89     streamable variant BCJ2 from 7-Zip could be implemented.
     90 
     91     Filter that autodetects specific data types in the input stream
     92     and applies appropriate filters for the corrects parts of the input.
     93     Perhaps combine this with the BCJ filter improvement point above.
     94 
     95     Long-range LZ77 method as a separate filter or as a new LZMA2
     96     match finder.
     97 
     98 
     99 Documentation
    100 -------------
    101 
    102     More tutorial programs are needed for liblzma.
    103 
    104     Document the LZMA1 and LZMA2 algorithms.
    105 
    106 
    107 Miscellaneous
    108 ------------
    109 
    110     Try to get the media type for .xz registered at IANA.
    111 
    112