Scott Putterman wrote:
>
> We used RDF originally because that was the architecture the mozilla
> project was going to use. The outliner we have now didn't exist and we
> didn't know it was possible to do it. We had ideas early on that RDF
> wouldn't scale for the thread pane, but we always figured that we could
> fix it and make it perform. We improved performance greatly (some of
> you may remember it taking 10 minutes to load a 5000 message folder) but
> never enough to get it to the level that we needed. The outliner widget
> is the first hope that we can achieve our performance goals.
I'd be very interested to find out where the major bottleneck is in the use
of RDF, since I'm considering a similar technology for a project I'm working
on. It's clear that it must impose some overhead on the whole scrolling
procedure, but can you give me some idea of what sort of operation the
code is spending all its time in? Before David Hyatt's idea, what sort of
schemes were you considering for meeting the performance targets, and do
you think they were totally unworkable?
Thanks,
Antranig.
