Features Download
From: Eugeni Dodonov <eugeni <at> mandriva.com>
Subject: Re: [Announcement] Next Mandriva release dates and schedule
Newsgroups: gmane.linux.mandrake.cooker.devel
Date: Saturday 13th November 2010 20:29:12 UTC (over 7 years ago)
On 11/13/2010 04:32 PM, katnatek-bdk wrote:
> Annual cycle is not new Mandriva did try and probe to do a bad idea for
> lake of news about Mandriva and for the ubuntu's phenomena.
> What makes Mandriva thing this time is good idea? 

Well, a couple of things.

First, we had the 2010.1 release which was in the version/release freeze
much longer than expected. This resulted in a an impressive time we had
to fix and nail down the final remaining issues and bugs, being able to
fix things instead of hurrying up with the deadlines and dates. It is
impossible to fix all bugs of course, but this extra time we had allowed
us to improve the distribution considerable.

Second, I believe that the 6-months release cycle everyone is doing
right now is crazy. It gives just some weeks of development time between
one release and another, and limited time to freeze and fix bugs, so the
developers must run against the clock and depend on the time after the
release to fix remaining or not that throughly tested issues with updates.

And third, I think that this will allow us to provide a more
LTS-supported distribution versions.

Having said all that, the cooker development can change as well I
believe. Instead of having a
"unstable->unstable->unstable->alpha->beta->rc->frozen->new distro"
cycle, we can have a more practical release development. Having a
long-term supported stable Distribution version, and a Development
version in parallel can fit both worlds. E.g.,:
 - the users who want to have a stable distribution, which does not
changes much over time, just works, packages are updated with fixes and
security updates, and is fully supported for years, can have it, without
having to worry that in 6 months their distribution will start to get
deprecated already.
 - users (like me, for instance) who prefer bleeding-edge stuff, even at
the cost of the stability, can use cooker and its periodic snapshots.

Of course, this will also require changes in:
 - how Main is supported
 - how cooker development is carried out
 - how can we guarantee the stability of the "stable" distribution
without loss the innovation brought with 6-months release cycle.

These things are being discussed and proposed internally, I promise that
I will announce more news soon, and - of course - will listen to yours
feedback on that.

Eugeni Dodonov
CD: 3ms