You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@wicket.apache.org by Martin Benda <be...@bendis.cz> on 2009/01/21 12:31:19 UTC

Re: WICKET-1917: a major performance issue

A small mistake in the subject - should be WICKET-1917 instead of
WICKET-1970 which is completely unrelated to this message... sorry!

Regards,
Bendis

Martin Benda píše v St 21. 01. 2009 v 12:08 +0100:
> Hi all!
> 
> The recently added feature WICKET-1917 "Automatically Load CSS files
> based on name matching" causes a major performance issues for one of our
> applications. The application constructs pages with hundreds of
> Components and for every single component the ResourceStreamLocator
> looks up the same CSS and JS resources over and over again.
> 
> Maybe, if the ResourceStreamLocator would cache the results, the
> performance impact would not be so high... But I would prefer if the
> framework would not perform such hidden and "automagical" functionality
> by default. A developer knows better when he wants to load a resource
> for his component.
> 
> Regards,
> Bendis
>

Re: WICKET-1917: a major performance issue

Posted by Peter Ertl <pe...@gmx.org>.
I personally prefer to leave out this kind of (black) magic at all.  
Also static html previews will be wrong.

Implementing language specific resource loading is no rocket science  
and can be easily rolled out on your own.

Only my 2 percent

Am 21.01.2009 um 12:31 schrieb Martin Benda:

> A small mistake in the subject - should be WICKET-1917 instead of
> WICKET-1970 which is completely unrelated to this message... sorry!
>
> Regards,
> Bendis
>
> Martin Benda píše v St 21. 01. 2009 v 12:08 +0100:
>> Hi all!
>>
>> The recently added feature WICKET-1917 "Automatically Load CSS files
>> based on name matching" causes a major performance issues for one  
>> of our
>> applications. The application constructs pages with hundreds of
>> Components and for every single component the ResourceStreamLocator
>> looks up the same CSS and JS resources over and over again.
>>
>> Maybe, if the ResourceStreamLocator would cache the results, the
>> performance impact would not be so high... But I would prefer if the
>> framework would not perform such hidden and "automagical"  
>> functionality
>> by default. A developer knows better when he wants to load a resource
>> for his component.
>>
>> Regards,
>> Bendis
>>