Speed, speed, speed

[Ted Leung <tw...@sauria.com>]

Okay,

Since I seem to be the only one whining about speed, I figured I'd just
bring
this up as a data point.  I was reading over in soap-user, see, and I saw
this
page referenced. http://www.extreme.indiana.edu/~aslom/exxp/

The bottom line:  we are getting pounded for small documents.   I think that
we need to try to address this in X2.   And this is another reason why I'm
getting the willies about some of the perf issues.  It doesn't really look
good
for us to get slower.

Oh, and those guys looking for a pull-parser use case can look here too.

Ted


---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-j-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xerces-j-dev-help@xml.apache.org

Re: Speed, speed, speed

[Arnaud Le Hors <le...@us.ibm.com>]

I'm not really surprised by these results, as far as I know nobody ever
worked on making the parser set and reset phases fast. Most performance
work, if not all, has been put into parsing a big file as fast as
possible.
Andy Heninger found the same thing in Xerces-C recently and brought it
to my attention. I meant to look into this for Xerces-J but haven't got
a chance. That's something anybody could pick up though! In particular a
first analysis would already be valuable. It's not like it's something
that requires deep knowledge of the internals of the parser.
-- 
Arnaud  Le Hors - IBM Cupertino, XML Strategy Group

---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-j-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xerces-j-dev-help@xml.apache.org