12:41 p.m.
Dear list,
I have problems to understand the current review process. Long story
short it seems that there might be an informal process which does not
match the one documented in [1]. Since I don't understand this informal
process I have problems how to handle the new package request [2].
With some more details: I have submitted the package and made a long and
in my eyes tough review. I'm aware that this package is controversial,
the very reason I submitted it was to clarify and find a way to handle
external repos within rpmfusion. After a thread on the,mailing list and
more than 50 comments the package was approved. My problem starts here.
After a (delayed) cvssync request it turns out that Nicholas Chauvet
requires me to do certain changes to actually perform the cvssync. He
does so in what he describes as his capacity as "package coordinator".
This is a problematic situation for me since some of the changes he
requires me to do voids the review discussion. I also find it strange
because in [1] there is no mention of any review activities once the
package is approved.
This is really about what to do when a review goes wrong (not saying
this has). In Fedora we have a well-documented process how to bring such
issues to the FPC or FESCO, but rpmfusion has seemingly no documented
process how to overrule a reviewer's decision (?) Without a working
steering group or similar instance these issues are hard to handle. In
particular, it's hard for me and Nicholas to find an agreement since we
have a different view of our roles.
One solution could certainly be to appoint Nicholas or someone else to
be the last resort to overrule reviews. Another would be to try to
breathe some life into the steering group. We could of course also try
to sweep it under the carpet. Whatever way we choose I think it should
be explicit and documented, hence this mail.
So: should we decide on some mechanism which is similar to FPC/FESCO
also for rpmfusion?
In the bug [2] there are some loose ends I would prefer not to discuss
in this thread. One is if I have been disrespectful to Nicholas. I'm of
course open to discuss also this, but then preferably in a separate
thread (or in private). Another thing is what changes should be done to
the package; this discussion belongs to the bug once the process is
clarified.
Cheers!
--alec
[1] http://rpmfusion.org/Contributors#Submitting_a_new_package
[2] https://bugzilla.rpmfusion.org/show_bug.cgi?id=3152#c54
6:48 p.m.
On Sex, 2014-06-13 at 14:41 +0200, Alec Leamas wrote:
In
particular, it's hard for me and Nicolas to find an agreement since
we
have a different view of our roles.
https://bugzilla.rpmfusion.org/show_bug.cgi?id=3152#c57
Nicolas, has arrived late for review, undoubtedly, could we think that
review is not approved yet (was approved by Richard) , and Nicolas is
the reviewer master ? .
For me thing is , make sense what Nicolas ask ? or you don't agree ? ,
if first, is simple please consider that review is not ended, if second,
"should we decide on some mechanism which is similar to FPC/FESCO " is
this list, RPMfusion are half a dozen of guys , that why we still in
cvs etc .
Best regards,
--
Sérgio M. B.
8:49 a.m.
New subject: rpmfusion review process
On 06/13/2014 08:48 PM, Sérgio Basto wrote:
On Sex, 2014-06-13 at 14:41 +0200, Alec Leamas wrote:
> In
> particular, it's hard for me and Nicolas to find an agreement since
> we
> have a different view of our roles.
https://bugzilla.rpmfusion.org/show_bug.cgi?id=3152#c57
Nicolas, has arrived late for review, undoubtedly, could we think that
review is not approved yet (was approved by Richard) , and Nicolas is
the reviewer master ? .
I suppose we could... but then this should be documented. And then
again, who has really the authority to update the process docs?
But if there should be some kind of "super-review" or whatever then we
need to formalize this so the submitter knows what's going on instead of
waiting for a seemingly overloaded cvssync administrator. Which is what
I first did, probably for a too long time. Note that this would make the
rpmfusion review process more complicated than Fedora's where there is
no further review actions once the package is approved, only
administrative tasks.
One alternative would be to say that also on rpmfusion the cvssync task
is an administrative task, not a "super-review" or so. To add a safety
net we could just state that anyone unhappy with a package (new or old)
could bring this to this list, effectively making the list take same
role as FPC/FESCO. IMHO, this would align the rpmfusion review process
better with Fedora's.
Of course, this only applies when there is a disagreement. Most remarks
are normally welcome be it under or after the review.
For me thing is , make sense what Nicolas ask ? or you don't
agree ? ,
Well, obviously we have a disagreement... I have tried to explain my
position in comment 58, which still is waiting for some kind of feedback.
if first, is simple please consider that review is not ended, if
second,
"should we decide on some mechanism which is similar to FPC/FESCO " is
this list,
Indeed.
Cheers!
--alec
1:05 p.m.
New subject: rpmfusion review process
Hi,
Sorry for late answer I'm completely overloaded
On Sáb, 2014-06-14 at 10:49 +0200, Alec Leamas wrote:
On 06/13/2014 08:48 PM, Sérgio Basto wrote:
> On Sex, 2014-06-13 at 14:41 +0200, Alec Leamas wrote:
>> In
>> particular, it's hard for me and Nicolas to find an agreement since
>> we
>> have a different view of our roles.
> https://bugzilla.rpmfusion.org/show_bug.cgi?id=3152#c57
>
> Nicolas, has arrived late for review, undoubtedly, could we think that
> review is not approved yet (was approved by Richard) , and Nicolas is
> the reviewer master ? .
I suppose we could... but then this should be documented. And then
again, who has really the authority to update the process docs?
We got a Steering committee http://rpmfusion.org/SteeringCommittee
Nicolas Chauvet was announced as master administrator by the previous master , some months
ago on this Mailing list but I can't find the email in question.
But if there should be some kind of "super-review" or
whatever then we
need to formalize this so the submitter knows what's going on instead of
waiting for a seemingly overloaded cvssync administrator. Which is what
I first did, probably for a too long time. Note that this would make the
rpmfusion review process more complicated than Fedora's where there is
no further review actions once the package is approved, only
administrative tasks.
One alternative would be to say that also on rpmfusion the cvssync task
is an administrative task, not a "super-review" or so. To add a safety
net we could just state that anyone unhappy with a package (new or old)
could bring this to this list, effectively making the list take same
role as FPC/FESCO. IMHO, this would align the rpmfusion review process
better with Fedora's.
we don't have any kind of FPC or FESCO because we don't have people to do that ,
so Nicolas is the FPC or FESCO .
Of course, this only applies when there is a disagreement. Most
remarks
are normally welcome be it under or after the review.
> For me thing is , make sense what Nicolas ask ? or you don't agree ? ,
Well, obviously we have a disagreement... I have tried to explain my
position in comment 58, which still is waiting for some kind of feedback.
Unfortunately, I don't had time to read it and I don't have time to moderate it to
see if parts can achieve to an agreement.
> if first, is simple please consider that review is not ended, if
second,
> "should we decide on some mechanism which is similar to FPC/FESCO " is
> this list,
Indeed.
Cheers!
--alec
Cheers!
--
Sérgio M. B.
3:08 p.m.
New subject: rpmfusion review process
On 07/15/2014 03:05 PM, Sérgio Basto wrote:
Hi,
Sorry for late answer I'm completely overloaded
Ha, I know that feeling :)
We got a Steering committee http://rpmfusion.org/SteeringCommittee
It's just that Hans, Thorsten and Matthias effectively has resigned. So.
there is no steering committee unless new members are appointed/elected.
Nicolas Chauvet was announced as master administrator by the previous master , some
months ago on this Mailing list but I can't find the email in question.
This is really good, we need someone who makes this job, and Nicholas
makes it really well. The problem is just that the master administrator
role isn't defined in any document (?).
In particular, is it the master administrator's task to make a "super
review" after the regular package review?
Cheers!
--alec
12:48 p.m.
New subject: rpmfusion review process
On 2014-06-13 20:48, Sérgio Basto wrote:
On Sex, 2014-06-13 at 14:41 +0200, Alec Leamas wrote:
> In
> particular, it's hard for me and Nicolas to find an agreement since
> we
> have a different view of our roles.
https://bugzilla.rpmfusion.org/show_bug.cgi?id=3152#c57
Nicolas, has arrived late for review, undoubtedly, could we think that
review is not approved yet (was approved by Richard) , and Nicolas is
the reviewer master ? .
For me thing is , make sense what Nicolas ask ? or you don't agree ? ,
if first, is simple please consider that review is not ended, if second,
"should we decide on some mechanism which is similar to FPC/FESCO " is
this list, RPMfusion are half a dozen of guys , that why we still in
cvs etc .
Best regards,
To add to my previous answer, not getting Nicolas view here on the
list
is making the review process even harder to understand.
I'm on a more or less disconnected holiday from now. If noone else
does, I'll try to resume this thread in late August.
Cheers!
--alec
3805
days inactive
3849
days old
rpmfusion-developers@lists.rpmfusion.org
5 comments
2 participants
participants (2)
-
Alec Leamas -
Sérgio Basto