[Fontconfig] Status about fontconfig CVS regarding cache handling

[Frederic Crozat]

Le mardi 31 janvier 2006 à 10:45 +0100, Frederic Crozat a écrit :
> Le lundi 30 janvier 2006 à 18:06 +0100, Frederic Crozat a écrit :
> > Le lundi 30 janvier 2006 à 11:40 -0500, Patrick Lam a écrit :
> > > Frederic Crozat wrote:
> > > > Hi,
> > > > 
> > > > I'd like to put a little status about the various cache bugs I've (and
> > > > other on the list) been encountering.
> > > > 
> > > > This is against CVS snapshot from today 2006-01-30 :
> > > > 
> > > > * fc-cache . bug => FIXED
> > > > * fc-cache /foo/bar vs fc-cache /foo/bar/ => FIXED
> > > > * fc-cat crashes on ~/.fonts.cache-2 file or on global cache files =>
> > > > FIXED
> > > > * ~/.fonts.cache-2 file being re-created when ~/.fonts exist each time a
> > > > fontconfig based application is still there.
> > > > * fc-cat ~/.fonts-cache-2 now returns empty information.
> > > > 
> > > > Once the ~/.fonts.cache-2 recreation bug is fixed, it would be probably
> > > > a good idea to push a 2.3.94 release so it is more tested by people not
> > > > living with CVS snapshots ;)
> > > 
> > > Thanks.  Let's see if current CVS is kind of stable and then I'll push 
> > > 2.3.94.  Also there's that make distcheck problem on my laptop (it works 
> > > fine on my desktop) which I don't know how to fix.
> > 
> > Status update with refreshed CVS snapshot (15min ago) :
> > 
> > ~/.fonts.cache-2 created each time an application is started when
> > ~/.fonts exist => FIXED
> > fc-cat ~/.fonts.cache-2 returns empty informations => FIXED
> > 
> > And I haven't see any regression about the other bug I mentioned in my
> > previous post.
> > 
> > As Matthias noted, "make check" now fails.
> 
> Status update with CVS snapshot 20050131 (10:44 CET) 
> * make check fails => FIXED
> 
> No regression found on my side.
> 
> I've just pushed it in Mandriva cooker to see if many people start
> screaming ;)

Well, just to let you know : cooker folks are screaming a lot :(

Many seems to have an old ~/.fonts.cache-2 file which get in the way to
force a entire cache regeneration for each application start.

I haven't been able to spot the problem precisely yet.

-- 
Frederic Crozat <fcrozat at mandriva.com>
Mandriva