Little changes in release schedule

[obgr_seneca]

After some discussion in the council and the packagers team, we decided yesterday to change the release schedule for Mga2 a bit.
This means adding a third beta and pushing the release dates of the rc and the final back.

You can read more in the mageia blog

Cheers,

Oliver

[ghmitch]

This news makes me EXTREMELY happy because it assures me that Mageia is obsessed with quality and not with pushing release frequency as were Mandriva. I absolutely hated, all through the late Mandrake and following Mandriva releases that their releases were dependably half-baked and guaranteed to break things and require many months of patches in order to make them fully functional again. After years of this, Mageia 1 has been a breath of fresh air. PLEASE ... do not take Mageia 1 away from us until you are certain that you have something better to offer with Mageia 2 AND, hopefully, a smooth and dependable update path as well. In the mean time, I will be happy to deal with the lack of a few cutting edge features in exchange for a solid and stable Mageia 2. - George

[doktor5000]

FWIW, nobody will take away Mageia 1, it is still supported until the end of the year.

[obgr_seneca]

Well, it's a two-edged sword.
As a community distribution it's much easier for us to push back release dares. There's no waiting custommers or cd printing companies and so on.
The other thing is, pushing a release date too much makes bad press as well. I think only Debian can do that

So we try to create a stable distribution, that doesn't have too many critical bugs (there'll always be bugs be it only because you don't have the hardware for testing everything...).
But we will also try not to push back the release dates too much...

Actually it's hard to make decisions like that, whatever you do, about 50% of the users, will hate you for it

Oliver

[ferri]

Some time I was running on Linux Mint.
Mint development leader has motto if I can paraphrase: It will be when it will be.
So new Mint release is when he is fully satisfied.
Due to this, there is not exact release date for Mint versions in advance.
And where is Mint? Distrowatch number 1.

But on other side I do not think that this is fully good approach because it brings a lot of unnecessary nervousness in community.

[doktor5000]

But on other side I do not think that this is fully good approach because it brings a lot of unnecessary nervousness in community.

If users are satisfied, because they know developers will not sacrifice quality just to keep up with some arbitrary release date, there's no need to be nervous.
But some users will always be notoriously nervous, where forum posts and mails will hit some days before release date, asking where the release is and so on ...

[ferri]

[quote="ferri"
But some users will always be notoriously nervous, where forum posts and mails will hit some days before release date, asking where the release is and so on ...

Yes, stupid people are everywhere.

[wobo]

But some users will always be notoriously nervous, where forum posts and mails will hit some days before release date, asking where the release is and so on ...

Yes, as always I'm looking forward to those, curious if any new combination of the usual words will turn up!

[andre]

And then there are users like me, that want a stable distro and also the latest packages. Running Mageia 2 beta 3, and firefox in the aurora chanel with some packages compiled because the ones in the repos were old, but mad when something crashes. Users like me are impossible to satisfy, but at least I know that the developer team can't make a distro exactly as I want it (I can't either, I'm not a developer. I don't even know how to package).

[ghmitch]

But on other side I do not think that this is fully good approach because it brings a lot of unnecessary nervousness in community.

What makes me nervous is the concern that the new product will be released before it is totally and completely debugged so as to be, if anything, BETTER than what I have now. That is one thing I HATED about Mandriva. Once it was reasonably debugged, you just new they would replace it with a new buggy version and cut support for the older one that finally worked (or at least sort of worked). Mageia 1 on the other hand is just about as bullet proof as you can get and it has been since it hit the ground. Mageia 2 had better be at least as good and hopefully better. - George

[ghmitch]

So we try to create a stable distribution, that doesn't have too many critical bugs (there'll always be bugs be it only because you don't have the hardware for testing everything...).

Of course, of course. But that can be remedied by maintaining support on the previous version UNTIL the new version is reasonably debugged AFTER it is released. In other words, the release dates should not be set in stone, but NEITHER SHOULD THE TERMINATION OF SUPPORT DATES. IF there are significant unforeseen bug issues in the new release POST RELEASE, support for the older version should be extended until those issues are resolved. - George

[wobo]

Of course, of course. But that can be remedied by maintaining support on the previous version UNTIL the new version is reasonably debugged AFTER it is released. In other words, the release dates should not be set in stone, but NEITHER SHOULD THE TERMINATION OF SUPPORT DATES. IF there are significant unforeseen bug issues in the new release POST RELEASE, support for the older version should be extended until those issues are resolved. - George

Did you see any talk or schedule where something to the contrary was said or written, to make you emphasize this point?

The primary goal is to release a version which has no (or next to none) release-critical bugs, that's why release dates are often postponed by distributors. Support for "release 1" will definitely go on for a while after "release 2" is out. But it is clear and usual all over the Linux world that this can not go on forever and a day. There has to be a point in time where the lifetime ends. As no software is totally bug-free there will definitely be bugs left unresolved when support of the previous release ends.

[ghmitch]

But it is clear and usual all over the Linux world that this can not go on forever and a day. There has to be a point in time where the lifetime ends. As no software is totally bug-free there will definitely be bugs left unresolved when support of the previous release ends.

I agree completely with both of those points. As I have said a number of times, I am absolutely delighted with Mageia. HOWEVER, I made the point because such was NOT the case with Mandriva and the last thing I would want to see is for Mageia to drift toward the Mandriva syndrome in this regard. - George

[wobo]

Well, all I've seen during the last 19 months show that if Mageia is drifting somewhere at all, it is not towards anything mandrivaish

[ghmitch]

Perhaps the biggest concern that I have regarding Mageia 2 is that the change to systemd goes flawlessly in an upgrade scenario. I have read several reports so far that would seem to indicate that it does and none that would indicate otherwise. I think it is important when doing a new release to undertake ONLY one major system architecture change and to MAKE SURE that the transition is transparent and that is no small order when you are dealing with a huge number of upgrades and due to infinite variations of hardware and software, no two are exactly alike. - George

[wobo]

Yes, that's why Mageia has started early (in Alpha stage) with those changes and issued many steps (Alpha, beta, RC) in between until the final, and at each step Mageia has asked for everybody to test.
Of course you can never test enough to exclude every possible hardware and/or software configuration there is. But hopefully the users of Mageia have understood the necessity to participate in testing.