Re: [PATCH 0/3] qom: Replace INTERFACE_CHECK with OBJECT_CHECK

[Eduardo Habkost]

On Thu, Sep 17, 2020 at 11:25:30AM +0400, Marc-André Lureau wrote:
> Hi
> 
> On Thu, Sep 17, 2020 at 11:19 AM Philippe Mathieu-Daudé <philmd@redhat.com>
> wrote:
> 
> > On 9/16/20 11:31 PM, Marc-André Lureau wrote:
> > > Hi
> > >
> > > On Wed, Sep 16, 2020 at 11:35 PM Eduardo Habkost <ehabkost@redhat.com
> > > <mailto:ehabkost@redhat.com>> wrote:
> > >
> > >     This series replaces INTERFACE_CHECK with OBJECT_CHECK because
> > >     both macros are exactly the same.
> > >
> > >     The last patch is a new run of the OBJECT_CHECK ->
> > >     DECLARE*_CHECKER* converter script that will convert the former
> > >     INTERFACE_CHECK-based macros.
> > >
> > >
> > > Well, at least having a different macro allows to tweak qom
> > > implementation or replace it with something different more easily.
> > >
> > > I have some wip branch somewhere where I actually made Interface a
> > > different beast than Object (it was saving some fields, and avoiding
> > > some potentially wrong casts iirc - I didn't bother to upstream that
> > > yet). Also I have a different branch where I played with GObject to
> > > replace qom. In both cases, your proposal would have, or would make, the
> > > work more complicated.
> >
> > Considering "wip branch not bothered to upstream" as "fork",
> > your comment from [*] applies here...
> >
> >   You can't blame upstream from doing cleanups and modernization, or
> >   stagnating. Forks are forks, with all the pain they carry. If they
> >   want to avoid the maintenance cost, they have to do the extra effort
> >   to get it upstream. This is also a "sneaky way" to remind them that
> >   effort is better spent in this direction.
> >
> > [*] https://www.mail-archive.com/qemu-devel@nongnu.org/msg720284.html
> >
> >
> Fair enough. Note I didn't nack it, but in general the proposal seems to
> blurry some worthy semantic difference between object & interface. Maybe
> #define alias INTERFACE_CHECK with OBJECT_CHECK instead ?

I'm actually considering deleting INTERFACE_CHECK and
OBJECT_CHECK completely after we finish the boilerplate
conversion, and tell everybody to use
OBJECT_DECLARE_INTERFACE/OBJECT_DECLARE_TYPE instead.

-- 
Eduardo