Am Donnerstag, 14. Juni 2012, 15:32:15 schrieb Dominique Leuenberger a.k.a DimStar:
I don't think this 'every big thing needs to go through staging' will help a lot to be honest. What we need more than ever is actually really people taking their packages serious! And communication about possible breakings. Like when we updated glib for example: communication upfront by the team updating it.. and actively supporting 'other packages failing due to that' (the factory status page helps a lot in giving an overview why something failed). But don't expect that one person can fix all the rest because he's updating one package... you might loose that one contributor taking care of a few packages and being willing to help fix others, if he feels he's *responsible* for all the other failings.
Though in a bit different way I think that a clear structure of responsibility is one of the keys for success – not only for Factory but every non-home repo. I have seen people updating repos and then not checking back (or ignoring) that it or parts of it fail to build because of their commit. Shoot and run. IMO there are two simple use cases. 1. packages like gcc which will break other packages and it's not the changed packages fault. In this case the responsibility to fix the broken packages is not on the submitter. The only thing he has to do is to notify about the change. Having a testing ground for that change before it goes to Factory would help to keep Factory clean. But it only works if people have to take over responsibility to fix their packages within a given time after they have been notified. How openSUSE enforces that is up to the organisation – and IMHO the crucial bit no matter what kind of plan you come up with. If you do not like it or are not able to enforce responsibility then you have to dedicate staff to cleaning up other people's mess, i.e. they take over the responsibility of others. 2. breakages due to packaging mistakes etc. People that commit (accept) changes to a repo should check back within the next 24h whether the repo still builds and if not fix their breakage. Weekend is not an excuse simply because if you are not willing to fix the breakage on the weekend that commit is not as important as to not hold it back until Monday. The question is again, how one can enforce people taking over responsibility. You can build some clever process on how things should work – yet unless you can ensure that people stick to it and take over the responsibility for their work it is not worth anything. In this context it would be helpful to have a switch in obs to only publish a repo if all packages for a distro version have built successfully. This would minimise the damage to users and keep the old, working packages. This does not catch bugs but only build errors. Sven -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org