> Update 2: Is there an actual need for a SPEC to ever use the %pkgbuilddir (or
> %builddir for that matter) macro? If not, why not make it at least "private"
> with an underscore?
Oh, I actually mulled about %builddir without the underscore as the
user-oriented name for this. I too kinda liked that more, I think I ended up
with the pkg-version to make minimize confusion with %_builddir. OTOH, like you
said we already have %buildroot and %builddir would seem more consistent with
that. I could quite easily be convinced either way :smile:
> Update 3: Oh... you actually mentioned that "a LOT of packages reference
> %_builddir for all sorts of (mostly bad) reasons" in the commit message. In
> that case, why introduce yet another such macro at all?
Primarily because %_builddir is ambiguous, you don't really know whether it's
the bad old or good new. It still points to the potentially shared
%{_topdir}/BUILD by default, whereas %pkgbuilddir (or %builddir) is only
defined during the actual build (which is also the only time that directory
exists).
Note that the "mostly bad" stops being bad with this change - this is now the
legit place for packages to use for their needs during the build.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2885#issuecomment-1970968229
You are receiving this because you are subscribed to this thread.
Message ID: <rpm-software-management/rpm/pull/2885/c1970968...@github.com>_______________________________________________
Rpm-maint mailing list
Rpm-maint@lists.rpm.org
http://lists.rpm.org/mailman/listinfo/rpm-maint
