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

Re: mono / ".pc files in -devel packges" guideline

On Sun, 2006-05-14 at 15:48 -0500, Tom 'spot' Callaway wrote:
> On Fri, 2006-05-12 at 18:27 +0200, Michael Schwendt wrote:
> > On Fri, 12 May 2006 08:59:29 -0500, Tom 'spot' Callaway wrote:
> > 
> > > On Thu, 2006-05-11 at 15:08 -0400, Brian Pepple wrote:
> > > 
> > > > > Even if it is for development, is it worth forcing it off into a
> > > > > -devel package for just one file?
> > > 
> > > If the .pc file is the ONLY file that would qualify for -devel (aka, no
> > > headers, no static libs, nothing to develop a program against), then
> > > there is no reason to force a -devel subpackage just for it.
> > 
> > There is one good reason: The .pc file contains dependencies on other .pc
> > files, which are located in -devel packages. In this case, you would need
> > a dependency ("Requires") from your non-devel package to -devel packages.
> > Without that dependency, your package would break the pkgconfig dependency
> > chain and several pkgconfig query commands.
> > 
> > To keep development files in -devel packages is the only clean solution.
> Somehow, this discussion spilled off the list, and I didn't notice that
> I wasn't replying here too.
> As Michael Schwendt pointed out above, pkgconfig files depend on other
> pkgconfig files (and in all released versions of FC, this is not
> automatically detected by RPM). This is a very good reason not to
> let .pc go into non -devel packages. If we permit .pc files in non-devel
> packages, then we'll have non-devel packages that depend on -devel
> packages, inflating the install needlessly. 
> I'd rather have one .pc file in a -devel package than trigger a chain of
> -devel packages to be installed when the user just wants to run a mono
> app.
> So, to be clear, I am not planning on changing the guidelines to deal
> with this issue. All .pc files need to go in -devel packages, even if it
> is the only -devel worthy file (which is the exception rather than the
> norm).

I'll modify the Mono Packaging page also.

Brian Pepple <bdpepple ameritech net>

gpg --keyserver pgp.mit.edu --recv-keys 810CC15E
BD5E 6F9E 8688 E668 8F5B  CBDE 326A E936 810C C15E

Attachment: signature.asc
Description: This is a digitally signed message part

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