Actually, I thought Johan pointed out what the problem was: that the
performance of MarkupContainer.add is something like O(n^2) with respect
to the number of direct children of the container, and that this was
part of a conscious decision to optimise for space rather than time.

Adding 100,000 children to the same container is therefore completely
unrepresentative of Wicket's real world performance.  A more realistic
benchmark would be to code up a more "normal" page and invoke it a
zillion times with JMeter or something.

jk

On Tue, Feb 19, 2008 at 01:01:49PM -0800, Dan Kaplan wrote:
> Why don't you use a profiler to find out?
> 
> -----Original Message-----
> From: Leucht, Axel [mailto:[EMAIL PROTECTED] 
> Sent: Tuesday, February 19, 2008 7:03 AM
> To: [email protected]
> Subject: AW: Performance
> 
> I deployed it in deployment mode but still same result. 
> 
> My guess is, that instantiating these lots of objects is taking its time.
> Can I somehow create just one label and while iterating manipiulate the data
> in the object?
> 
> /Axel
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to