|
News
News
- 23 May, 2004
- This is to inform you that xq-cpp has now been released as a file
download. Please visit the release page
to download it. Release
Notes. Preliminary benchmarking results show that xqML files are
atleast 2x as small as equivalent XML files and parsing speed is 4
times as slow as expat, with most of the time spent in looking up
symbol translations. This can easily be improved in a performance
optimized version. Compare this with other
slow performers for a better idea.
- 16 January, 2004
- New design and code of Xqueeze Reference Implementation in C++
is now in CVS. Check out xq-cpp module for the latest.
Last working draft of the specifications is now a release, with some important
changes. Expect a C++ release and benchmarking results soon.
- 19 September, 2003
- A new Working Draft of the Xqueeze specification has been
posted. Most important changes include:
- Optional encoding without the knowledge of schema
- Support for 8-bit and 16-bit character encoding streams
- Improved encoding rules
Get it from the Documentation page.
- 18 August, 2003
- Work on the Xqueeze project had been very low-key since April 2003
but starting September 2003, we're going to start development of a
complete xqML parsing suite and get some benchmarking done to show
that it works. We've also updated the specs to now allow generation
and parsing of documents without the knowledge of schema, which should
greatly help in scenarios where a lot of vocabulary mixing happens and
different vocabularies are processed by different pieces of software.
Meanwhile, if you want to contribute to the project, do
drop in a mail to tnhashmi AT
sourceforge.net. In particular, we'd be interested in people who
know Perl, Python, Ruby etc. and are comfortable with processing
octet-streams in these languages.
|
Project Links:
|