On Tue, 2014-06-17 at 20:01 -0600, Stephen John Smoogen wrote:
On 17 June 2014 18:35, Adam Williamson <awilliam(a)redhat.com>
wrote:
> On Tue, 2014-06-17 at 17:48 -0600, Stephen John Smoogen wrote:
> > On 17 June 2014 16:02, Adam Williamson <awilliam(a)redhat.com> wrote:
> >
> > > I've revised the release criteria draft again, with reference to the
> > > useful discussions both on-list and at this morning's meeting:
> > >
> > >
>
https://fedoraproject.org/wiki/User:Adamwill/Draft_server_release_criteria
> > >
> > > I added the firewall exception for the Cockpit web interface, clarified
> > > the issue about role deployment "at install time", and added
new
> > > criteria for the cockpit management interface to be running OOTB and
> for
> > > roles to meet their "functional requirements, as defined in their
role
> > > specification documents" - role specification documents being
something
> > > I invented out of my ass at the meeting this morning. View that one as
> a
> > > trial balloon. :)
> > >
> > > As always, thoughts / comments welcome!
> > >
> >
> > OK. First of all, where could I 'test' any of these things on a Fedora
20
> > system.
>
> Well, that would be pointless. We're building a new product, called
> Fedora Server. That's kind of the whole point. We already have release
> criteria that are more or less scope-appropriate for the product called
> "Fedora 20" - they're the Fedora 20 Release Criteria, which we used
to
> validate the Fedora 20 release.
>
>
That is not what I meant. I am sorry I am not communicating well and not
being helpful here. I have seen the links and such but they are in the
words on a white board. There are ~60 days before the alpha+2 weeks and I
wanted to see what code was written and possibly set it up against Fed 20
(as Fedora 21/Rawhide may not be the best to test against as its changing)
so I could see if the draft looked spot on or if it was too little or too
much. That was all.
Again my apologies
Sorry for sounding harsh, I was just a bit frustrated. So, let me try
again: if you have concerns about our ability to implement the Server
tech specs within the current F21 timeframe, well, that's a thing. The
Server WG actually has similar concerns, which is why it asked FESCo for
a schedule modification recently, but that's not exactly how it came
out.
The point from *my* perspective, though, is this: the place to raise any
such concerns is probably not in relation to the release criteria. I
think I've been consistent in believing that you don't do product design
via release criteria; the criteria are built *from* the product design,
not vice versa. Initial product design for Server was all done some time
ago, and all done very publicly and openly - see all those links I cited
in my first post. All I'm attempting to do with the criteria draft
process is come up with appropriate release criteria for the product
called "Fedora Server" as it is currently designed. The question of
whether we'll actually be able to implement the design within the
proposed timeframe is an important one, but it doesn't seem helpful to
try and deal with it in the context of the release criteria.
If anything, it's important that we have the release criteria and, more
generally, all the other bits that go with it - a validation test plan,
basically - *in order to know whether we're actually succeeding in
that*. If you want to look at it this way, you could say that what we're
building here is the mechanism by which we will figure out if we've
actually managed to build the thing we designed.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net