On Sunday 15 April 2012 12:32:35 Andrew Benton wrote:
> On Sun, 15 Apr 2012 11:43:41 +0100
> 
> Ragnar Thomsen <[email protected]> wrote:
> > On Saturday 14 April 2012 22:57:32 Pierre Labastie wrote:
> > > - vlc has several required (or at least recommended) dependencies, that
> > > is, there should be either a configure switch --disable-package or the
> > > 
> > > package should be installed, specifically:
> > >  > either libmad installed or --disable-mad passed
> > >  > either ffmpeg installed or --disable-avcodec passed
> > >  > either liba52 installed or --disable-a52 passed
> > 
> > I fixed the dependencies. I chose to set libmad, liba52 and ffmpeg as
> > required deps for VLC, since (in constrast to Lua) most people would want
> > these.
> Wouldn't it be more accurate to have them a recommended deps? The book
> is about giving people information, if you say the deps are required it
> makes it seem that they're required, which is not true. If you make
> them recommended then people know that you expect the deps to be
> installed but if they choose to do it differently they can.
> 
> I know I'm a hypocrite for saying the above as there are a few packages
> where I've put pkg-config as a required dep but I know they could be
> built without it if I set loads of CFLAGS and stuff. It's a judgement
> call, where do we draw the line?

Well, for VLC, the configuration fails if these packages are not present or 
the respective --disable-* switches not used. So if we set them as recommended 
deps, I guess we should have the disable switches in the instructions? Or 
would it suffice to provide them in the command explanation section?

I think it's difficult to draw the line. I guess the best strategy would be to 
list all deps that can be disabled with simple configuration switches as 
recommended/optional. If more advanced deviations are needed, such as 
modifying CFLAGS etc they should be required deps.

-Ragnar-

-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Reply via email to