best practices for updates in stable releases

[Pavel Alexeev (aka Pahan-Hubbitus)]

Patrice Dumas wrote:
> There are some places where some interaction happens, and some checks of
> guidelines, are possible, like in reviews. Also some specific cases
> have policies (missing maintainer, who is allowed to change other people 
> packages). But in general maintainers are free to do whatever they want
> in their packages.
> 
> In fact I think that it is a good thing, except when bugs are not fixed,
> even though the bug reporter provided with a fix.
Great words!
Free is freedom is freedom with all pluses and minuses! Freedom it is 
not only be free doing want you want, it is also allow do it for others!

Sure, please do not threat it as proclamation to anarchy!

We have guidelines, and I believe what most of maintainers tried be 
closer to it. We pedantic check following to it in step where it may be 
formalized: In review request, in reviewing, in sponsoring, in 
guidelines change, in many other situation. But we can't even try apply 
any sanctions to maintainer to its treat some very common 
recommendations like [1]

Sorry, but it is only recommendations. F.e. how you are suggest apply 
any sanction by formulation like (cite [1]): "New upstream releases 
should not necessarily be pushed to release branches. The benefit of the 
bugfixes and new features should be weighed up against the risk of 
regressions.". How it should be weighed? Who should weight it? For what 
scale in last??? Must we call FESCO on each update to vote it? Do karma 
absolutely necessarily (so, some rigidity and qualification here even I 
wish see, but it is subject another thread)? Any things?

[1] 
http://fedoraproject.org/wiki/PackageMaintainers/Package_update_guidelines