Mailinglist Archive: opensuse-project (229 mails)

[Stanislav Visnovsky <visnov@xxxxxxx>]

On piatok 19 November 2010 17:32:12 Thomas Schmidt wrote:
> On 19.11.2010 11:06, Stanislav Visnovsky wrote:
> > Hi!
> >
> > On utorok 16 November 2010 16:59:28 Andreas Jaeger wrote:
> > > On Tuesday 16 November 2010 13:33:18 Stanislav Visnovsky wrote:
> > > > Hi!
> > > >
> > > > On piatok 12 November 2010 07:06:54 Satoru Matsumoto wrote:
> > > > > After yesterday's openFATE screening team IRC meeting, I've overviewed
> > > > > recently requested features (via new instance) and got some additional
> > > > > impressions and questions which should be shared and further
> > > > > discussed.
> > > > >
> > > > > # You can read the mention below on http://openfate.titanpad.com/notes
> > > > > # as well. ;-)
> > > >
> > > > I was thinking about the issue of postponing features. I would propose
> > > > to use the 'Package Wishlist' approach, i.e. use backlog
> > > >
> > > > The idea is to use collect all features that are not ready to be
> > > > integrated into a release (underdefined, no good solution found, ...)
> > > > in a backlog bag. Only when we have an idea, how to implement a
> > > > feature, ideally to have already someone willing to work on the
> > > > feature, the feature would be marked for particular release.
> > > >
> > > > This approach will eliminate postponing features forever, at least for
> > > > big number of features where this would happen. Also, it would keep us
> > > > from cleaning up huge list of features for each release.
> > > >
> > > > Did we consider this approach before? If yes, what are the reasons
> > > > against it?
> > >
> > > I have been considering as well a general "openSUSE distribution"
> > > product but was not sure how to do it properly. Could you elaborate a
> > > bit more when this backlog will be used and what will be done with e.g.
> > > openSUSE 11.4?
> >
> > Right now, most of the features are not really specific for release. They
> > are ideas that might be good to be implemented. As we don't know better,
> > right now we assign them to the next openSUSE release and keep that 
> > postponing for the ones that did not get done (majority).
> >
> > So, I would introduce an "incubator" for ideas - where even a short
> > description is enough. From here, the openFate screening team picks them
> > up, works on filling details up to the point where the idea is
> > sufficently understood to be implemented (status "Marketplace").At this
> > point, we need to find someone willing to implement the thing - and if
> > there is someone, this feature can be finally moved to particular
> > release (because only here more or less know the timeframe when it's
> > going to be ready).
> >
> > It's similar to agile approach, but much better reflects the reality how
> > we do openSUSE IMO.
> >
> > There is a question of the 'picks for next release' - those 5 mandatory
> > features we want to focus on. For them, we might want to assign the
> > release already in the Marketplace state.
> >
> > I don't have a good answer if to use Products for this or States.
> > Products seems to be more natural choice given the way openFate works
> > right now. Might be the case the current openFate does not fit this
> > proposal.
> >
> > I would like to hear your opinions, if this is direction makes sense to
> > others.
>
> I think it makes sense to collect the features in this incubator product.
> At the moment we need to reject and move features to the next product on
> and on... We had such a bucket product some time ago, and it got removed.
> But I think that was because it did not make sense for business products.
> But we should also have a procedure that features do not rot forever in
> the incubator product which also makes a bad impression.

I don't think it's a problem if a feature stays in Incubator for long time. 
This should not happen for features that get votes though.

Stano
-- 
To unsubscribe, e-mail: opensuse-project+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse-project+help@xxxxxxxxxxxx