On 5 May 2015 at 21:40, Martin Schlander <martin.schlander@gmail.com> wrote:
It was also the first release based on the "new" factory/tumbleweed, with a very short testing phase (two devel releases with 2-3 weeks between) and iirc the roadmap itself was published very late.
Secondly, openSUSE releases shouldn't need as many testers/much testing as before, with Tumbleweed already being fairly well tested by a decent number of people running it full-time. At least I thought that was the idea.
The whole process of Tumbleweed based stable releases just needs a few small adjustments, to work well imo. (A roadmap, a slightly longer window to test things, a fixed 12 month release cycle, 26 months of lifetime (2 releases+2 months) would be a very good start).
From an Engineering perspective it was 'just' a Tumbleweed snapshot, but from a Marketing, Branding, 'actually getting a release out of the door' perspective, it required as much work as any other software release. For 13.2 this fell on the shoulders of a tiny handful of people. The reality is that Tumbleweed has captured the excitement of a huge
I understand where you're coming from, but I think there are a few things I'd like you to consider The 13.2 release required a large amount of work from a small number of people to make it a success. portion of our current contributing community, and grown it greatly, and that's great, but it does have the side effect of leaving less enthusiasm and less people to work on the Regular Release. As was discussed at the Project Meeting at oSC 15, the 'Tumbleweed-Snapshot-based-Release' model also has an interesting side effect when it comes to this part of getting a release published and marketed. Just as they share development models, both Tumbleweed and snapshot-based-releases have very similar core messages - "The latest of everything, each release". The major differences are the gap between releases and the method of maintenance after release. On each Release Day for the Regular Release, this ends up leaving us with relatively little exciting to talk about (Everything is already in Tumbleweed) and at the same time, way too much to talk about (the Changelog of 13.1 to 13.2 is insanely long and hard to filter down into exciting things to include in Release announcements) If snapshot-based-releases were to continue, we'd need to find a way to tackle that, and get people excited, talking about, and helping put together the release and it's marketing, something which is basically just the same as what we do every week, but maintained differently. But this is where my proposal for a Regular Release based on the SLE Sources comes into play. ( Video available at https://www.youtube.com/watch?v=BH99TSrfvq0 ) Our Regular Release has always been pulled in two directions, with some wanting it faster and more innovative, some wanting it slower and more stable. Tumbleweed has addressed the 'go faster' crowd - I think we have the *best* Rolling Distribution out there, and it's getting better Now we have the sources for SUSE Linux Enterprise we have an opportunity to build the *best* Stable Distribution out there. Something that really addresses the needs of those users who want a stable Linux distribution for 'real work', with Enterprise stability at it's core, with the community turning it into something exceptional. The best of both worlds, which is really something different from what everyone else has (Ubuntu LTS, CentOS) - and it's steps in that direction which will help us get new people involved to help work on it and make it even better going forward. I realise this proposal would mean the change of expectations for each openSUSE regular release - maybe we wouldn't upgrade every version of [insert software here] each release. I realise this is a huge change from the promise of 'The latest of everything, each release'. The individual teams of our project will have the opportunity to choose what goes into the new Regular Release and how often they want to upgrade it. With the clear division in direction between Tumbleweed and this new Regular Release, hopefully decisions about which versions make sense for inclusion when become easier for those teams. They'll know the kind of expectations our users have for either distribution, and be able to make informed judgement calls as to when to include whatever they want from the various upstream projects they maintain in openSUSE. In the case of the software provided in the shared SLE Sources, I also think we also have a lot of collective learning to do, as I think many of the fears regarding the 'age' of that software and their capabilities don't hold up under scrutiny (eg. the SLE Kernel gets a TON of backported features and patches) So while, of course snapshot-based-releases is a valid option, I'd rather see what we can make of this opportunity to build something better for openSUSE Project and our users. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org