[maemo-developers] [maemo-developers] Re: Taste the Herring!

[From: David Weinehall]

On tor, 2006-11-30 at 12:30 +0100, ext Sebastian Spaeth wrote:
> David Weinehall wrote:
> > When did Linus last announce the release date of the next kernel (and
> > actually follow it)?  Ohhh, but I guess you don't regard that as an open
> > and community participated effort either. =)
> 
> OT, but not such a good analogy :-).
> 
> Linux has to spent quite some time elaborating on whether he should 
> release in 2-week or 2-year cycles, if there is to be a parallel 
> development and stable branch and when/how often branching is supposed 
> to happen. He also has to justify and make clear whether he will put out 
> security fixes for released versions or not.

Well, the thing is, he never follows his own plans.  If that's the same
behaviour you expect, sure, we can say that "we'll release at least
bi-yearly", then release week 1, week 49, week 52.  Because that's
pretty much how the kernel release-cycle looks like.  Linus releases
when he thinks things are ready.  Personally I find that nice.
If I want a new feature earlier than that, I'll use a git-snapshot.  
Similarly, if you want a Maemo-feature earlier than the next release,
use herring or sardine.

The only real deadline Linus follows is the initial merge window.  The
one (or is it 2?) week(s) when he accept new features (as opposed to bug
fixes).  And that kind of deadline wouldn't make sense for us, since
that's an internal policy.

> The topic on when he actually followed his own plans is a different 
> topic :-).

No, it really isn't, not in this case.  Because the important thing
isn't the plan, but the execution of the plan.  A plan that's never ever
followed isn't a useful plan.

The security fixes, stable branches, etc taking place in the 2.6-tree
are all volunteer efforts from other than Linus.  If you want the same
for Maemo, by all means, go ahead and start making Maemo 2.1-spaeth1.
Maybe you'll even get a following =)


Regards: David