Matthias Wimmer wrote:

Hi!

[EMAIL PROTECTED] schrieb am 2003-09-04 22:20:21:


Additionally, I don't think really performance is an issue (at least with
the transports) right now. When I started working on AIM-t, there were
several errors in it that kept it crashing even in small environments. Now
there are only several memory leaks left ;-).



I know that Java can be fast and I don't want take part at the flames on different languages.

But one fact I want to add: It's not that the transports are just small
processes and the Jabber server does the most. At amessage _most_
computing power is needed for the transports. This may be because they
are coded so badly or because they use libraries that were never
indended to be used for transports. But nobody should think that you
don't have to care for performance in transports as they don't have to
do much, this is just not true.


Tot kijk
Matthias


I think it's code style or bad implementations, really.
I.e. i used in enigma3 (a client most don't know) a library called tinyXML. Yes, the library is tiny. :-)
But, the tiny library stored the complete XML tree in memory - i think what happened doesn't have to be said ... heh ... and there was no way for me to find a way how to change the root node.


well, i ended up writing my own XML parser which allowed me my things i needed to have.

last thing was pcurtis fixed a bug in my XML parser :)
i think there is no way for an individual to write bug free code - no way.

ulrich

--
Ulrich B. Staudinger
http://www.die-horde.de
email: [EMAIL PROTECTED]
jid: [EMAIL PROTECTED]

current project: REDHORN
http://redhorn.sourceforge.net


_______________________________________________ jdev mailing list [EMAIL PROTECTED] http://mailman.jabber.org/listinfo/jdev

Reply via email to