release schedule proposal

[Inge Wallin]

On Wednesday, February 02, 2011 11:26:10 Boudewijn Rempt wrote:
> On Wednesday 02 February 2011, Inge Wallin wrote:
> > I may have misunderstood, but if I understood correctly I'm very sorry to
> > have to say that I hate this. The reason is that it is a
> > developer-centric way to work rather than a user-centric.
> > 
> > I'm sure it will be very ego-boosting for us developers to get new
> > versions with new fancy features out there as soon as possible.
> > 
> > But as a user, I expect to be able to update to either a new version with
> > new features and possibly also new bugs *or* a more stable release with
> > no new features but also definitely less bugs.
> > 
> > If this is the way that we will do things, then the user will live in a
> > state of perpetually changing set of bugs and more or less well-working
> > new features.
> > 
> > For a short time, until we reach something like a big enough feature set
> > that a real organization can use it, then perhaps this could be a way
> > forward.  But I know that admins at larger installations prefer
> > stability much before new features.  And stability is what will suffer
> > with this scheme.
> 
> But we don't have that sort of users at all -- there are no admins with big
> installations of koffice or calligra.

While this is true, the Linux distros work the same way. I think we should ask 
the packagers first what they think of the idea.

> We don't have the kind of users who
> value stability over experimentation, since there hasn't been an end-user
> release of KOffice since 2007. The only app with something approaching a
> userbase is Krita, and there the users are all for having the
> possibilities they ask for in their hands as soon as possible.

> What we do have is an opportunity to show that there is progress because we
> are in a situation where master is no longer unstable like trunk used to
> be, but can be kept quite stable, while still receiving new features on a
> regular basis.
> 
> And in the end, having frequent snapshot releases _is_ a stopgap until we
> reach the release scheme Cyrille proposed: four releases a year, of which
> one gets bug fixes for a longer period.

Good!  I didn't see this in the original suggestion, though.  Hence my 
objections.  I dislike it a little bit less now. :-)  Do you have any thoughts 
on when to interrupt this scheme and go to X releases per year (4 is not yet 
decided)?