Re: [Nautilus-list] [PATCH]: avoid re-layouting when it's not needed.



Yoann Vandoorselaere <yoann mandrakesoft com> writes:

> Maciej Stachowiak <mjs noisehavoc org> writes:
> 
> > On 17Jun2001 12:58PM (+0200), Yoann Vandoorselaere wrote:
> > > 
> > > On directory load, we do relayout several time before it's needed
> > > (before any file get added or changed). In fact, in this case, the
> > > files_changed signal should only be hooked after finish_loading is
> > > called (when the view is loaded and we are starting monitoring the
> > > directory), and disconnected as soon as the directory load is done.
> > > 
> > > This work fine on my local tree, and seem to help with latency.
> > > 
> > 
> > Darin said he was worried about this change introducing bugs. Can you
> > describe what testing you did to make sure it doesn't, 
> 
> First can you comfirm me that a files_changed signal can and should
> be caught by the directory view between two directory load...
> 
> I know the files_added one is used, not sure about files_changed...
> 
> But I might have a better fix soon that would avoid making this risky
> change... I'll keep you informed...

Ok, this second fix might exist (do not process a NautilusFile if it
contain a request) but still wouldn't be better than avoiding the 
signalling (As pointed out earlier GTK signal are slow).

So the real question is (and I really need an answer to this one) :
Should fm-directory-view caught signal after a directory was loaded,
or is it only usefull while loading the directory ?

(I checked and couldn't find a way where the directory view
 use a files_changed signal after the directory loading).

-- 
Yoann Vandoorselaere | Tiniest "mesures unities?"
MandrakeSoft         | - lenght : millimeter
                     | - volume : milliliter
                     | - intelligence : military man




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]