[Openstack-devel] OpenStack summit April 2012 post mortem

[Loic Dachary]

On 04/20/2012 01:04 PM, Thomas Goirand wrote:
> On 04/20/2012 02:37 PM, Loic Dachary wrote:
>> Essex is our target release for Wheezy and we can rely on the fact that it will be actively / passively maintained for a year.
> Unfortunately, we need it to be maintained for 3 or 4 years. Let's say
> if Wheezy is out next February, then we have 2 years of Wheezy being out
> stable, and one year being old-stable.
>
> Did you highlight that Debian needs this commitment? 
Dave Walker (Ubuntu) and myself advocated that they need to be maintained as long as there is someone to backport the security fixes.
> Are Ubuntu people
> willing to do security and bug fixing together with Debian for this
> period, since Ubuntu LTS will be released with Essex? 
It is my understanding. They would not commit to maintain all releases for 5 years and therefore did not commit to this. In the proposed cycle, the Essex branch will be volunteer maintained ( and no commitment from the OpenStack project, as opposed to the commitment to actively maintain / passively maintain ) and because Ubuntu has an Essex based LTS we can assume they will publish security fixes.

One might fear that a corporate backed distribution such as Ubuntu does this stable maintenance behind closed doors or even in a framework that is separate from the project. I think it's a good sign that Dave Walker insisted on keeping this long term stable branch maintenance within the realm of the OpenStack project, as opposed to an Ubuntu specific repository.
> Is there anyone
> else willing to help for this maintenance?
A number of companies are in a bad place because they put a lot of effort into Diablo and did not contribute back. As a result it's now very complicated for them to upgrade to Essex and even more complicated give back what they've done because there is an enormous difference. Some of them, such as HP, decided to start this difficult reconciliation work. Others will find ways to deal with the problem. In any case, there now is ample proof that the cost of going alone for stable branch release is not worth the gain.

This is the kind of experience that a project needs to discover by itself it seems, because the problem is hardly new and there are many ways to deal with it.

So the answer to your question is : we don't know yet but I'm reasonably optimistic about the fact that the stable branch maintenance will greatly improve starting with Essex.

Cheers