Re: [SPAM] [nebula-dev] [CompositeTable] very bad scalability

[Tom Schindl <listom@xxxxxxxxxxxxxxx>]

Hi,

Well I think VIRTUAL Support is making it's way slow but it's making 
it's way. The problematic part currently are Viewers working with 
DeferredContent-Providers because this is not working as designed and 
Boris even once mentionned that he thinks we need to deprecate the whole 
package. Still as I discovered most of the time the Datamodel is most of 
the time not the limiting factor when working with Tables, it's the 
creation of all the TableItems in none VIRTUAL mode.

I've used normal content providers with VIRTUAL-Tables successfully 
because I'm always fetching the all the data once and don't query my 
datasource all the time this way the table-items are created when needed 
but the buggy behaviour of the deferred content providers is worked around.

Tom

André Dietisheim schrieb:
> Hi Tom
>
> It's the wrong list but a very short answer you help a lot:
> are all bugs related to virtual tableViewer gone in the current
> cvs-code?
>
> Regards
> André
>
> On Mon, 2007-03-12 at 13:48 +0100, Tom Schindl wrote:
>   
> > André Dietisheim schrieb:
> >     
> > > Hi Tom
> > >
> > >       
> > > > The current implementation provides the possibility to completely use
> > > > tables/trees without any mouse interaction. I even have patched for me
> > > > locally combo-editors to directly open the DropDown for users when they
> > > > click in and the same is possible for editors using Dialogs.
> > > >         
> > > sounds great! Is SWT performance good enough to instantiate widgets on
> > > traversal?
> > >       
> > Well the Widgets are not instantiated every time they are only
> > hidden/shown :-) so the performance is good enough. Take a look at the
> > snippets available in
> > http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.jface.snippets/
> >
> > Tom
> >
> > _______________________________________________
> > nebula-dev mailing list
> > nebula-dev@xxxxxxxxxxx
> > https://dev.eclipse.org/mailman/listinfo/nebula-dev
> >
> >
> > !DSPAM:45f54c4c56131327716623!
> >
> >     
>
> _______________________________________________
> nebula-dev mailing list
> nebula-dev@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/nebula-dev
>