Open Stack

I tend to agree with Lucas. I think ironic-lib was designed to have common
code to be used primarily into ironic and IPA. It was not intended to be
used in other projects. We tend to add methods in the ironic-lib with
Ironic and IPA in mind and never thought of it could be used by out of tree
drivers.

But it would be useful, it were to consumed by out of tree drivers.

Thanks,
Shiv

On Mon, May 16, 2016 at 8:26 PM, Sam Betts (sambetts) <sambetts at cisco.com>
wrote:

> I personally disagree with saying that if we wanted it make it usable by
> projects other than ones in the Ironic umbrella it should go into oslo. I
> think that non-ironic projects directly related to Ironic such as out of
> tree drivers etc, should be able to utilise the code placed into
> ironic-lib.
>
> Neutron are doing a very similar thing for all their drivers/extensions
> they have broken out over the last 2 cycles,
> http://specs.openstack.org/openstack/neutron-specs/specs/liberty/neutron-li
> b.html.
>
> Making ironic-lib available to out of tree drivers etc also puts us into a
> good position to begin the work to stabilise things like the driver API.
> Neutron is making the rule that out of tree drivers shouldn¹t
> inherit/import anything from the neutron core code base, only neutron-lib,
> they are doing this to provide a stable interface that shouldn¹t be broken
> by changes to neutron core. I think we could do the same, with in-tree
> drivers dog-fooding the driver api we provide in ironic-lib.
>
> Sam
>
> On 16/05/2016 15:14, "Lucas Alvares Gomes" <lucasagomes at gmail.com> wrote:
>
> >Hi,
> >
> >Thanks for starting this discussion Ruby.
> >
> >On Mon, May 16, 2016 at 2:57 PM, Loo, Ruby <ruby.loo at intel.com> wrote:
> >> Hi,
> >>
> >> A patch to ironic-lib made me wonder about what is our supported usage
> >>of
> >> ironic-lib. Or even the intent/scope of it. This patch changes a method,
> >> Œbootable¹ parameter is removed and Œboot_flag¹ parameter is added [1].
> >>
> >> If this library/method is used by some out-of-tree thing (or even some
> >> in-tree but outside of ironic), this will be a breaking change. If this
> >> library is meant to be internal to ironic program itself, and to e.g.
> >>only
> >> be used by ironic and IPA, then that is different. I was under the
> >> impression that it was a library and meant to be used by whatever, no
> >> restrictions on what that whatever was. It would be WAY easier if we
> >>limited
> >> this for usage by only a few specified projects.
> >>
> >> What do people think?
> >>
> >
> >I still believe that the ironic-lib project was designed to share code
> >between the Ironic projects _only_. Otherwise, if it the code was
> >supposed to be shared across multiple projects we should have put it
> >in oslo instead.
> >
> >The governance description is a bit vague [0], it does say: "A python
> >library of common ironic utilities.". Does it include out-of-tree
> >ironic-related code? I would like to think that it does not. The fact
> >that ( IIRC ) we never stated that this is a public library, there's
> >no documentation of any methods/modules there, no release notes, no
> >sample usages, etc... Makes it a bit unsuitable for 3rd party
> >consumption.
> >
> >That said, I'm not totally against make this library suitable for a
> >broader usage but this will require more thoughts on it and effort.
> >
> >[0]
> >
> https://github.com/openstack-infra/project-config/blob/920441f2fc02df89624
> >5b53e6f18b4ea6a0252a0/gerrit/projects.yaml#L2193
> >
> >Cheers,
> >Lucas
> >
> >__________________________________________________________________________
> >OpenStack Development Mailing List (not for usage questions)
> >Unsubscribe:
> OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> >http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>
>
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20160516/81ab8c1d/attachment-0001.html>