XML Compression Tools / News: Recent posts

Encoding bug

Version 0.98.1 (and previous versions) had a serious bug in the encoding for files with a large number of element or attribute symbol names (for example, large, heterogeneous files) that makes decompression impossible. This bug is fixed in 0.98.2.

Earlier versions are still available so that existing files can be decompressed.

Posted by James Cheney 2004-11-26

Version 0.98.1 and Binary Release

Version 0.98.1 is up, along with win32 and linux binary versions. Slight changes were necessary to get version 0.98 to compile correctly under Windows and Cygwin.

Posted by James Cheney 2004-06-10

Version 0.97 recall

Version 0.97 has several serious flaws, and will be deactivated. Users interested in stability and UTF-8 compression should use version 0.96; users interested in compressing non-UTF-8 documents or improved compression performance should use version 0.98.

Posted by James Cheney 2004-06-08

Version 0.98 source release

XMLPPM version 0.98 is now available in source form. It still needs to be tested under Windows/cygwin, but runs fine under Linux.

Version 0.98 corrects a serious bug in the encoding that was introduced in 0.97; I don't know what I was thinking. In addition, 0.98 incorporates several minor improvements and fixes. In addition, decompression speed has been improved by buffering and blocking calls to libiconv.

Posted by James Cheney 2004-06-08

Version 0.97 source release

Version 0.97 is ready. Improvements include

* faster, better compression
* support for compressing and decompressing XML files in all character encodings supported by libiconv

Posted by James Cheney 2003-02-24

Updated Windows binaries

The expat library name fhanged to libexpat.dll on Win32. So I recompiled everything; the new binaries are in win32 version 0.96.1.

Posted by James Cheney 2002-08-02

XML compression survey

I have put a survey up concerning XML compression. I am curious to see what the perceived needs for compression in various XML application areas are, and I'm hoping to have get a sense of what to do next.

Posted by James Cheney 2001-06-14

Windows binaries available

I have placed the Win32 versions of xmlppm online. These also require the Windows version of expat; in particular the expat DLL is needed to run them.

Posted by James Cheney 2001-05-23

Linux binaries up

I put two packages containing compiled, optimized Linux versions of the old and new versions. Windows versions coming soon.

Posted by James Cheney 2001-05-22

Version 0.96 released

Version 0.96 of xmlppm is now available in source form. (I will put some binaries up soon.)

This is a major upgrade. Xmlppm 0.96 uses Dmitry Shkarin's PPM implementation, called PPMDG. PPMDG is much faster and somewhat more effective than the underlying PPM compressor used in earlier versions; as a result, xmlppm 0.96 is *faster* than bzip2, within a factor of 2 of the speed of gzip, and compresses better than version 0.95. Of course, files compressed by earlier versions are no longer compatible.... read more

Posted by James Cheney 2001-05-19

Mailing list created

I made a mailing list, xmlppm-users.

Feel free to join and discuss the fascinating world of XML compression. No, really, I'd like to hear about what people are interested in using this stuff for.


Posted by James Cheney 2001-05-11

Now runs under Win32

I fixed a couple of trivial bugs, so now XMLPPM runs properly under Windows NT (and presumably other flavors).


Posted by James Cheney 2001-05-11

Initial release

XMLPPM 0.95 has been released. XMLPPM is an XML-specific compressor based on the PPM text compression algorithm. While rather slow, XMLPPM can compress XML files considerably (5-20%) better than existing text or XML compressors such as gzip and XMILL.

XMLPPM is in alpha, so don't use it for anything critical.

Contributions to developing XMLPPM are welcome. Since I'm a somewhat busy graduate student and this isn't my main area, it's unlikely that I'll have time to do much development on this, so I'd particularly appreciate additional interested developers.... read more

Posted by James Cheney 2001-02-09