https://bugzilla.rpmfusion.org/show_bug.cgi?id=2300
--- Comment #25 from Alec Leamas <leamas.alec(a)gmail.com> 2012-06-27 12:15:03 CEST
---
(In reply to comment #24)
Basically I think this kind of library, containing some files intended to be
included by other packages at build time, is also a bundled library in the
sense of the packaging guidelines. You could compare e. g., all of the
header-only boost packages; there's a lot of them. There's a discussion on a
similar scenario in
https://bugzilla.rpmfusion.org/show_bug.cgi?id=2140.
Another thing is that it's not intended to produce so-file(s) to be linked with
an application. That makes the idea to provide only a -devel pkg the right one
IMHO.
As for the pycvs discussion: whatever happened then this is a new situation:
there are now at least two packages using pycxx. So the reasons to unbundle is
stronger now. Also, I get the impression that the community is somewhat more
strict on this today than in 2008.
One possible approach might be to apply for an exception until the pycxx
package is approved; John needs to be sponsored and that could possibly take
some time. If the package is configured to pick up an existing system copy that
should acceptable IMHO. Hm... maybe that's the way to go, decoupling freecad
from pycxx?
We should preferably make some tests with pycxx package installed.
But let's focus on the rest for the moment, basically the filtering which is
what's blocks a complete review ATM.
--
Configure bugmail:
https://bugzilla.rpmfusion.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.