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

Re: file dependencies and packages and [blocker] bugs



On Fri, 2008-02-29 at 16:18 -0500, seth vidal wrote:
> On Fri, 2008-02-29 at 20:03 +0100, Hans de Goede wrote:
> 
> > Okay, these are mine (which I already knew without looking at the report, as I 
> > very deliberately added the filedeps):
> > 
> > /usr/share/fonts/bitstream-vera/Vera.ttf
> > xblast-0:2.10.4-5.fc9.i386
> > chess-0:1.0-12.fc9.i386
> > 
> > /usr/share/fonts/dejavu/DejaVuSans-BoldOblique.ttf
> > sdljava-demo-0:0.9.1-9.fc9.i386
> > 
> > /usr/share/fonts/dejavu/DejaVuSans-Bold.ttf
> > sdljava-demo-0:0.9.1-9.fc9.i386
> > 
> > /usr/share/fonts/dejavu/DejaVuSans-Oblique.ttf
> > sdljava-demo-0:0.9.1-9.fc9.i386
> > 
> > /usr/share/fonts/dejavu/DejaVuSans.ttf
> > sdljava-demo-0:0.9.1-9.fc9.i386
> > 
> > 
> > The problem with these packages is that the application in question (== 
> > game-ish app, using its own fontrending) opens these fonts by filename. not 
> > true any known font configuration mechanism.
I am having the same issue with packages I maintain.


> Define 'regularly' for the fonts changing locations?
For me, the problem had been fonts having vanished and fonts having
changed location _unnoticed_ between FC7 and FC8 (Initial FC8 contained
broken packages, because of this)

As a consequence of this, I am hard-coding deps on font files. My
packages now break when fonts vanish/move.

>  Also - If they are
> changing so often why don't we talk to the font maintainers and get that
> to stop moving around?
They changed due to conventions on font installation dirs changed and
because fonts where dropped due to licensing issues.

> > On the postive site, very few people have either of these installed, so I guess 
> > the best solution here is to notfix. But I'm open for other ways of catching 
> > filename changes in another package without using file deps.
> 
> It makes things like repoclosure take longer to run b/c it has to go get
> the filelists to resolve it all out.
Well, it would be a mistake not check them - These deps are present,
whether you like it or not.

> Don't you think we should be able to do something to make this situation
> better or even rarer?
I think you are wasting your time on a broken heuristic.

Ralf



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