On Tue, Aug 09, 2022 at 01:56:43PM +0200, Pavel Hrdina wrote:
On Tue, Aug 09, 2022 at 12:54:09PM +0100, Daniel P. Berrangé wrote:
> On Tue, Aug 09, 2022 at 01:52:33PM +0200, Pavel Hrdina wrote:
> > On Mon, Aug 08, 2022 at 01:22:17PM -0400, Daniel P. Berrangé wrote:
> > > One specfile containing both native and mingw builds is the
> > > new best practice for Fedora. This reduces the maint burden
> > > and ensures the mingw packages don't fall behind.
> > >
> > > Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com>
> > > ---
> > > .gitlab-ci.yml | 2 +-
> > > libvirt.spec.in | 287 ++++++++++++++++++++++++++++++++++++
> > > meson.build | 17 +--
> > > mingw-libvirt.spec.in | 327 ------------------------------------------
> > > 4 files changed, 293 insertions(+), 340 deletions(-)
> > > delete mode 100644 mingw-libvirt.spec.in
> >
> > The patch looks good but there are some changes not mentioned directly.
> >
> > With this patch we will build MinGW packages by default on Fedora. Not
> > sure if that is desirable. I would rather have it the other way around
> > if it works for Fedora best practice.
>
> Fedora has shipped the native & mingw builds for years now. This just
> merges them into one spec. There's no change in what we actually build
> from Fedora POV. Or am I misunderstanding what you mean ?
From Fedora POV everything is probably the same but for everybody else
this might be regression that would require using the --define as we
need to do for gitlab-ci.
IMHO if a contributor using the upstream spec to build RPMs, they need
to just deal with whatever the current packaging has defined. Ultimately
you can still do 'dnf builddep' to get the list of deps installeds,
including the mingw ones now.
With regards,
Daniel
--
|:
https://berrange.com -o-
https://www.flickr.com/photos/dberrange :|
|:
https://libvirt.org -o-
https://fstop138.berrange.com :|
|:
https://entangle-photo.org -o-
https://www.instagram.com/dberrange :|