[Libosinfo] [osinfo-db/libosinfo PATCH 00/18] Deal with "deprecated" and "removed" devices
Daniel P. Berrangé
berrange at redhat.com
Mon Nov 12 09:56:58 UTC 2018
On Mon, Nov 12, 2018 at 10:47:44AM +0100, Fabiano Fidêncio wrote:
> On Mon, 2018-11-12 at 09:31 +0000, Daniel P. Berrangé wrote:
> > On Thu, Nov 08, 2018 at 07:55:19AM +0100, Fabiano Fidêncio wrote:
> > > This patchseries aim to have OSes properly dealing with deprecated
> > > and
> > > removed devices. The path I've taken is:
> >
> > The notion of "removed" devices makes sense as a break to stop the
> > inheritance from earlier OS version.
> >
> > I'm more interested in the notion of deprecated devices though. Can
> > you give a clear example of where that would be used. To me this
> > feels
> > like an overly crude/coarse concept and so not convinced we should
> > have
> > that.
>
> I have to admit that I don't have a concret example that could be used
> right now. However, what I had in mind while working on those patches
> was the case of a, for instance, cpu.
>
> Let me try to explain ... *if* we have the "pc" type as part of osinfo-
> db, we could just mark it as "deprecated" and apps would have a quite
> easy way to check that instead of having to check different devices in
> order to decide what to use (as done right now with q35).
I don't think that's a good use of "deprecated". "pc" type is not
deprecated from the POV of guest operating system, it is from the
POV of the hypervisor. A guest OS will happily support one or the
other, or both. The mgmt app should decide which is preferred based
on its broader view of the system.
> As this is *not* something that we'd have a practical use soon, I'm
> more than fine on just dropping the patches related to the "deprecated"
> notion.
>
> Do you want a v2 with the patches dropped?
Yes please.
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 :|
More information about the Libosinfo
mailing list