RFC: kde stability proposal (redux)

[Roderick Johnstone]

On 09/03/10 01:11, Kevin Kofler wrote:
> Rex Dieter wrote:
>> One more call for discussion around another proposal that came out of
>> FUDCon
>> Toronto.  Short version of this is that we'd consider slowing down updates
>> a step, esp for the second half of a fedora release's lifetime, and to
>> limit kde 4.x-type upgrades to at most 1 per release.
>>
>> http://fedoraproject.org/wiki/SIGs/KDE/Stability_Proposal
>
> Considering the results of Adam Williamson's poll of Fedora users and
> considering how I still don't see a reason why we'd treat the previous
> stable release any differently than the current stable (there is no such
> policy anywhere in Fedora, it's just pointless second-class treatment), I'm
> still very much against this proposal.

Although I'm not sure how representative of the whole community the poll 
is, I'd like to strongly agree with Kevin here. In the research group I 
support we usually upgrade to every other Fedora release, and having the 
  previous to current release less well supported in the KDE department 
would be poor for us.

>
> If the update is good enough to be stable for F12, why is it not good enough
> for F11 as well? As for workload, the things which are the most work
> (bumping the specfile, removing obsolete patches, occasionally fixing file
> lists, making sure it builds) have to be done anyway, building the same
> stuff one more time is basically trivial anyway.

Many comments in various threads seem to have picked up on the 'reduce 
the workload for the KDE team' idea. The current system works very well, 
it minimizes the work for the KDE team, and keeps all supported Fedora 
release as well supported as possible given the available effort. I dont 
see a need to change it.

>
> I really see no benefit whatsoever in not upgrading the previous stable
> Fedora, and instead I see major drawbacks (no more bugfixes, except for
> possibly select few backported ones (but there's no way we can backport all
> of them!); additional workload for us because we can no longer just sync the
> specfile to fix issues; etc.).

Agreed.

What I think some people don't like is any change to the desktop or new 
bugs getting introduced with major new features. However Fedora is known 
and has as one of its underpinning principles to 'lead' rather than 
'follow', so I'd hope that just being rather more explicit in the 
documentation (somewhere) about what the updates system for KDE is would 
go a very long way to resolve this issue. After all the KDE team are 
fantastically responsive to issues.

In summary, the system as it is seems to work very well for us, and it 
looks to me as if it also minimizes the work for the KDE group.

If I were to be able to vote, I'd vote to keep the updates system as it is.

Roderick