June stable push ?

Thorsten Leemhuis fedora at leemhuis.info
Sat Jun 13 10:10:45 UTC 2009

On 12.06.2009 18:24, Kevin Fenzi wrote:
> On Fri, 12 Jun 2009 14:00:06 +0200
> Thorsten Leemhuis <fedora at leemhuis.info> wrote:
>> On 09.06.2009 21:44, Xavier Bachelot wrote:
>>> Is anyone taking care of the stable push for June ?
>> FWIW, there IIRC wasn't even one in May either iirc.
>>> I guess Thorsten is probably busy with other stuff,
>> Yes, sorry, I was really busy with lot of RPM Fusion stuff and real
>> life; EPEL moved off by radar so I forgot about it.
>> In fact I feared exactly that and that was the reason why I months ago
>> (is it even more then a year now? not sure, can't remember) tried to
>> get away from EPEL nearly completely (¹), which was running quite
>> fine back then.
>> But the latter not the case anymore. Sure, packagers are still doing a
>> whole lot of good work, but leadership/steering is afaics nearly not
>> existent at all, which IMHO is a dangerous situation for a project
>> like EPEL (especially as RHEL6 is not that far away anymore and we
>> need to be prepared for that).
> Well, somewhat agreed. What critical issue is waiting to be addressed?

I didn't follow the EPEL that closely, but these are the things that
sprung to my mind without consulting the archives:

- the "java is now in RHEL and EPEL" issue
- how to prevent things like the java issue in the future
- prepare for RHEL6 -- how do we get lot of Fedora packages into EPEL
for release day, as it later get hard for packagers
- can we make peace with CentOS and Dag somehow?
- can we make support CentOS in the phases better, where RHEL X.Y is
out, but Centos ist still on X.(Y-1)
- koji/bodhi status
- do we have one look and feel? I got the impression that some packagers
update their packagers more in a Fedora-like way, while others are more
debian-like (which up to a point is okay, but I'm not sure if we have
left that point behind us)

There are likely more things.

>> Some examples: No weekly reports for months (which were a requirement
>> from FESCo when the EPEL Steering Committee was formed and thus should
>> still be written!), nearly no meetings and the steering committee
>> obviously doesn't even care if testing -> stable move get done. Hey,
>> not even one of the steering committee members actually answered your
>> mail in the past few days which IMHO tells us everything already.
> I have tried (twice or three times, I forget) to get regular meetings
> going again. Either I can't get a time where everyone can attend, no
> one does attend, or people forget about the meetings. 

Meetings itself are not important, solving problems and improving the
project is -- that can be done via the list to if people want to.

>> Or IOW: what exactly did the steering committee do over the past six
>> months? Nearly nothing I'd say. Dglimore (maybe one or two others from
>> the Steering Committee; not sure,) did some work for koji and bodhi
>> for EPEL, but that's nearly all afaics.
> Agreed. What items do you think should be addressed? 

See above.

>>> iirc, there are some perms issue.
>> That is still the case afaics. Only dglimore can actually do the push
>> and testing->stable moves.
> Yes, this is the case as far as I know. 

Another problem that IMHO should have get solved -- relying on one
person and putting all the push-work on that persons shoulder IMHO is
bad for everyone.

>> (¹) Yes, I did prepare a few stable pushes over the past few months to
>> help out, but I mentioned a few times already that I didn't really
>> want to do them anymore; looking back at it I should have made it
>> more clear to prevent the current situation; sorry for that;
> yeah, sorry if people have been bugging you on that. 

No need to say sorry. Actually I would have had not much of a problem to
prepare another push or two if the Steering Committee or someone would
have asked me to. It just felt of my radar -- and even if I had
remembered it it likely would have felt a bit like "hey, seems the
Steering Committee is mostly inactive, so why should I invest my time then".


More information about the epel-devel-list mailing list