To freeze or not to freeze?

[Andreas Pakulat]

On 17.01.10 03:03:10, Aleix Pol wrote:
> On Sun, Jan 17, 2010 at 2:51 AM, Andreas Pakulat <apaku at gmx.de> wrote:
> > On 17.01.10 00:56:47, Aleix Pol wrote:
> > > On Sat, Jan 16, 2010 at 11:12 PM, Andreas Pakulat <apaku at gmx.de> wrote:
> > > > Do you have something that you want to implement? So far it seems
> > nobody
> > > > really has any feature he thinks needs to be done for 4.0. And making
> > it
> > > > official that we're in freeze means we have all the good reasons to
> > > > concentrate on bugfixing instead of playing around on a new feature
> > > > which might not even be finished for 4.0.
> > > >
> > > > Well there's a leap between adding features and being freezed. It's
> > kind of
> > > hard to fix stuff when you can't change strings, for example.
> >
> > Not at all, there are tons of bugs that don't need change of strings.
> > But as I said, we can probably delay the string freeze a bit.
> 
> But we want to do free and liberal bugfixing during the meeting :)

We can probably also do lots of bugfixing with a feature freeze in
place. 

> > > Plus if we came up with any feature why not add it?
> >
> > Because we actually want to release KDevelop 4.0. We need to stop adding
> > features to do that and concentrate on getting the outstanding bugs
> > fixed.
> >
> > > > > We don't have a final release date, do we?
> > > >
> > > > David suggested end of march back when we decided to slip KDE 4.4
> > > > release date and nobody objected.
> > > >
> > > How much time do we want to be freezed (in RC I assume) until we
> > released?
> > > We could build a release schedule from the end of the meeting.
> >
> > Well, we had the option of having 2 months of freeze and it was decided
> > thats not enough. As the amount of big changes seems indeed to slowly
> > drop I think doing a 2 month freeze would be enough then. And that means
> > freezing before the meeting, unless we want to delay the release yet
> > another month.
> >
> > So far in all the mails where I've asked nobody had any problems
> > releasing end of march (which in turn means freezing start of Feb at the
> > latest). We really need to find some conclusion on the dates and then
> > stop changing things around just because someone had a good idea for X
> > or Y. Else we can just stop worrying alltogether and not release at all.
> >
> > Andreas
> 
> Well, the meeting is in mid february, we could release in early april.

We need to find an end to the delays.

> Having RC's from right after the meeting is over.

That doesn't work. A proper RC needs at least a week of stabilizing if
there's no freeze during the meeting. We already now do at least 3 days
of freeze before each beta and that definetly is needed to have a
working beta (we had 2 cases where we didn't have long-enough freezes
with rather bad beta's).

Anyway, I'm going to stop argumenting into either direction. So far
David is for an early freeze. Apparently you and Milian are for freezing
after the meeting. And I'm undecided, so you two win unless more people
raise their hand for a freeze.

Andreas

-- 
You will have a long and boring life.