Bugzilla searches went down from "so-so" to "unusable"

Brennan Ashton bashton at brennanashton.com
Sun Aug 10 17:10:20 UTC 2008


On Sun, 2008-08-10 at 17:34 +0200, Michael Schwendt wrote:
> On Sun, 10 Aug 2008 09:29:30 -0500, Jerry Amundson wrote:
> 
> > On 8/10/08, Michael Schwendt wrote:
> > > On Sat, 09 Aug 2008 21:21:07 -0430, Patrick O'Callaghan wrote:
> > >
> > >> > > > With the older bugzilla, one also encountered those time-outs when
> > >> > > > using machines with less than 2 GHz clock rates. With the new
> > >> > > > bugzilla, the requirement for processor power has increased a lot.
> > >> > >
> > >> > > Processor power in the *client*?
> > >> >
> > >> > Yes.
> > >>
> > >> It's a web interface to a database. Is the database bigger or more
> > >> complex than Expedia or Travelocity or Amazon? Does the interaction
> > >> require more complex Javascript than Google Docs? I don't get it.
> > >
> > > It's not the database size. It's the complexity of the dynamic search
> > > form. Multiple thousands of package names per several product versions,
> > > for example.
> > 
> > I'll venture to guess that Expedia and Travelocity and Amazon have
> > complex searches. ;-)
> 
> You don't understand. I don't refer the searching, but the web forms
> and the client-side processing power that is needed to build them
> dynamically with Ajax.
> 
> > There's more to it. Non-optimized code or SQL? Hardware? OS tuning?
> 
> We're talking past eachother. The problems with bugzilla are on the
> _client_.
> 
It seems that it would make more sense in this situation to have a auto
complete box rather then a list box. That would keep the client very
minimal at a small server time expense.




More information about the fedora-test-list mailing list