Re: timeout patch for properties - take 2

[Alexander Larsson <alexl redhat com>]

On Thu, 2003-09-18 at 09:21, Mathieu Lacage wrote:
> On Wed, 2003-09-17 at 16:55, Alexander Larsson wrote:
> > 
> > This is used for the initial refresh everything, but it means it does
> > the whole properties_window_update() even if only
> > updated_deep_count_in_progress is needed.
> > I think not refreshing if files == NULL, and instead passing TRUE in a
> > dirty_all argument when doing the initial properties_window_update()
> > would work better. To avoid calling
> > directory_contents_value_field_update() unnecessary there should be a
> > directory_contents_dirty flag too, which get set by the 
> > "updated_deep_count_in_progress" signal handler.
> 
> I don't think this optimization is worth it: the code as-is works really
> pretty well. If someone complains about performance, maybe we can fix it
> later. If you don't agree, I can add that to the patch though...

Maybe not. Although deep_count_in_progress is gonna keep being called
for a while, especially for directories with deep trees in them, can you
at least avoid doing the file-change updates for that? Thats quite a
small change.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 Alexander Larsson                                            Red Hat, Inc 
                   alexl redhat com    alla lysator liu se 
He's a scarfaced zombie romance novelist fleeing from a secret government 
programme. She's a scantily clad hypochondriac museum curator who dreams of 
becoming Elvis. They fight crime!