From: Frank Gevaerts <frank_at_gevaerts.be>
Date: Mon, 25 Aug 2008 19:20:30 +0200

Hi all,

We should have a supported release out pretty soon now. That means we
will actually have a decent build for people to install (that is what
a release is after all).

Currently it is more or less expected that every commit improves rockbox
(except for accidents), and that no change gets committed that has known
instabilities or introduces new known bugs.

Maybe we should relax that a bit for the first few weeks after the
release. This should not impact "normal" users as those would be
expected to run 3.0, and it would allow much wider testing for
patches that right now don't get applied because of a too great
perceived risk.

After a few weeks (maybe a month or so?) we would then go back to our
regular "every revision should work well" regime.

This of course also has some implcations for how we handle support for
the release itself, as it won't necessarily be as easy anymore to ask
people to try the current build.

I don't want to propose timelines and rules right now. I think we should
first reach a consensus on the basic principles. So do we want something
like this, or should we remain as conservative as we are now at all times?

Frank

-- 
"Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it." - Brian W. Kernighan

Received on 2008-08-25