#1173 provenpackager request
Closed None Opened 6 years ago by vicodan.

Hi FESCo. I'd like to join the provenpackager group.

Justification:

I am reluctant and hesitant to file this ticket, but I wouldn't ask if I didn't need it.

I am currently working with on Enlighenment 1.7.8.

https://fedoraproject.org/wiki/Changes/Enlightenment

Ideally I'd like to get the packages in to F19 but I'm experiencing unforseen circumstances.

https://bugzilla.redhat.com/show_bug.cgi?id=1003692
https://bugzilla.redhat.com/show_bug.cgi?id=891244
https://bugzilla.redhat.com/show_bug.cgi?id=998774

Compounding the RHBZ's noted above (small sample) in addition to lacking acls for packages like evas-generic-headers, libeio, libeina, and others I cannot proceed further.

History:

I'm the primary maintainer for MATE desktop and the mate-compiz spin, Co-maintainer for Cinnamon (and hopefully future [co]maintainer of the spin).

I currently own/maintain/pay attention to 113 pacakges.

https://admin.fedoraproject.org/pkgdb/users/packages/vicodan

FAS: vicodan
IRC: dan408

I appreciate the consideration. Thanks in advance.

Dan


It's not FESCO who decides, but a jury of your peers :)

+1 from me

Not a vote, just a note in order to disambiguate the various Dan's we have. His real name is Dan Marshal, Fedora page here:
https://fedoraproject.org/wiki/User:Vicodan

That seems a limited set of packages, did you request ACLs ? If you plan to maintain E17, you'll have to deal with them regularly anyway.

-1

Some may call me bad, but I'm not confirmed that Dan knows what he's doing.

Acquiring commitacl access to the packages would be the road to take. Even lots of packagersponsors need to do that nowadays, because else they could not touch their sponsoree's packages.

Plus, communicate better. That also refers to the other package maintainers, who may not respond to pkgdb requests.

I also don't understand the "unforseen circumstances" that have been referred to.

The "libeio.so.1" is a result of both the submitter and the reviewer not having done the review painstakingly enough. Recently, there have been other packages that have been rushed through the review process. I couldn't believe my eyes, when I saw the comments at
https://bugzilla.redhat.com/show_bug.cgi?id=1003692#c3

For "evas_generic_loaders vs. evas-generic-loaders", I think also the FPC is to blame (they haven't decided on the opened naming guidelines clarification ticket yet), since in the guidelines there is confusion in several places about what the "upstream name" is. If it's the upstream project name, one would construct a package name from "Evas Generic Loaders" -> evas-generic-loaders (as has been reviewed and approved before), but if one is expected to prefer the tarball name, it would be evas_generic_loaders as in the later review request. This could have been noticed during review, however.

Further, bug 1003196 has been run into recently, too. Subpackage file duplication madness in mate-utils. That's really not how we do packaging.

-1

I recall few sad incidents in the past when Dan was involved (from technical point of view or from social/community pov). I don't want him wearing provenpackager badge.

Moving to FESCo meeting item per policy on negative votes.

In response to the -1 votes because of some negative communication in question on the devel list (over a year ago), this was with my now comaintainer of MATE, Wolfgang and has now been resolved. There was also ONE other minor one, but again I'd like to stress I've come a long way.

I have worked extremely hard to repair the relationship I have with Wolfgang and we are now best of friends.

I have been heavily mentored by rdieter.

I intend to adhere to all provenpackager guidelines including not touching any packages I shouldn't be touching.

I would like to remind everyone that I have nothing but the best intentions here and that I didn't know that becoming a proven packager required "knowing everything".

I would also like to state that I have heavily read the proven packager requirements and have consulted with people before applying.

Thanks,
Dan

full disclosure, being dan's sponsor I'm probably totally biased, but... :)

Dan has come a long way over the course of the past 3-6 months (ie, if you'd asked me back then, I'd probably have said -1). He's become a good collaborator and either knows what's best already or at least who and how to ask for advice. I also like the his level of mentoring and reviewing lately (hint: becoming a sponsor is with considering).

+1, I think dan is ready, and would use these powers for the proper good.

AGREED: vicodan's request for provenpackager was approved (+6,-0,0)

-1

I do not agree to this decision and the way it was made.

There are two problems here:

  1. The request itself:
    I don't think Dan is ready for proven packager yet. I have been filing bugs against his packages recently and his replies and commits indicated that he did not understand stuff that was documented well in the packaging guidelines.
    I cannot confirm that the MATE maintainers are "best friends", in fact, my perception is very different. MATE in Fedora suffered from it's two maintainers not getting along too well. Dan has been pushing development versions into the repos without discussing the implications with Wolfgang. As a result, MATE in F20 will be a mix off different versions which are not considered ready for release by upstream.

  2. The Process:
    If the policy includes a period of one week for the sponsors to cast their vote, the request should not been taken to FESCo before that week is over.
    And if several sponsors state concerns, FESCo should not simply overrule them but take them into account. Looking at the meeting According to [http://meetbot.fedoraproject.org/teams/fesco/fesco.2013-09-11-18.00.log.html#l-93 the meeting logs] I think FESCo members only relied on Rex'es judgement even though he said he is clearly biased.

I would like to ask all FESCO members who voted who of them actually looked at bugs or commits of the requester. If you haven't, you should not overpower the sponsors who have.

P.S.: Dan, please don't take any if this personally. It's not so much about you but about the MATE maintainers (that is you ''and'' Wolfgang) and about FESCo. I'm frustrated about the process in first place. I think you can be proven packager, but you are not ready yet.

I share cwickert POV on that matter and kindly ask fesco to revise its decision.
Provenpackager is not only about technical skills but also social skills. Based on previous incidents -i would also consider checking mailing lists-, there's a room for improvement in the latter domain for dan.

And since there was two minus votes (moreover from two respected maintainers), i think that fesco should have either declined or postponed its decision.

cwickert could you add relevant bugzillas? I don't plan to go over all open mate bugs (30) and all commits.

Personally, I was +1 for his request, because his mentor and sponsor spoke about improvements. Also his emails on fedora-devel seems to be ok.

I guess for revert of our decision there must be some evidence.

I am personally -1 to revert unless there is evidence that he used his new provenpackager powers incorrectly and harmfully.

Replying to [comment:12 cwickert]:

-1, from me to this, too, for the following reasons:

  1. I fully agree with cwickert's pov about errors in the process here.

  2. There's only two people working on MATE. Why should there be a need for being a provenpackager? Dan can ask / collaborate with Wolfgang, if he really needs something.

  3. In the Enlightenment-team there's already a provenpackager (sundaram), who can do needed fixes / rebuilds / buildroot-overrides for Dan, if he is in need.

  4. Dan could additionally ask Rex Dieter (his sponsor) if he is in need for a provenpackager, too.

  5. Wolfgang has created a [https://git.fedorahosted.org/cgit/comps.git/commit/?id=b4611be58a54990ea10ecf2be7acb2ce22f30725 nice and slim group for MATE in comps] and without any discussion or notice Dan [https://git.fedorahosted.org/cgit/comps.git/commit/?id=720de223d126ca85fc3d4693d87f043d0d28de88 kicked it out two weeks later] replacing it with a 1200+ packages "monster", because "a seperate group is too complicated and duplicate". For which reason?

  6. I'm absolutely not fine with someone, not being able to collaborate or '''discuss changes''' with the maintainer / author '''prior to doing them''', to have access to my packages.

Dear FESCo-members, please re-think your decision.

-1 from me as well, and I agree with cwickert that the decision here was rushed without a good reason before 7 days had passed.

I personally haven't collaborated a lot with Dan, and when I have, I have actually been able to work fine with him without having conflicts. A lot of other people however haven't.

For instance, https://bugzilla.redhat.com/show_bug.cgi?id=972372#c5 was one of the tickets where I strongly disagree with Dan's handling of the issue. Dan grabbed a core GNOME package when it was briefly orphaned and then proceeded to do changes that would have left a number of other GNOME packages in a FTBFS state, right before the F19 release. And then didn't respond to mclasen's tries to contact him on IRC.

nirik had a hour long argument with Dan on IRC after which he reverted that change.

On another occasion, Dan comments:
"Have YOU ever tried working with them [GNOME upstream]?? It's like pulling teeth.

Please get over it."

https://fedorahosted.org/fpc/ticket/305

In my opinion, a provenpackager should be someone who can work with other people, and not try to overrule others at every possible occasion.

P.S. I would fully support Wolfgang Ulbrich's (the other MATE maintainer) provenpackager request if he wants to become one. He has been doing a lot of Fedora related work recently and knows how to collaborate with other people.

@mmslano: I though that the following law was pretty much spread among Fedora contributors "Any discussion that involves Dan and GNOME is bound to end badly"

https://lists.fedoraproject.org/pipermail/devel/2013-January/177328.html <= purposely rude to Mairin Duffy instead of exposing his point.

https://lists.fedoraproject.org/pipermail/devel/2013-July/186867.html <= more recently, could have expressed his opinion without being rude.

I think that Dan is a decent guy and a good contributor, but he has to be more tactful toward his peers. Being considerate to other contributors is a strong pre-requisite to any responsability position (provenpackager included) within Fedora.

I think we should not judge persons or their personality here. I can be a stubborn idiot, too. I think we all are from time to time. And frankly speaking I wouldn't call these examples particularly rude. I know many other people who could have written the same or worse - and still be excellent proven packagers.

What we need however judge is are the persons ability to fulfill his duties, and for proven packagers that includes both the technical skills and the will to collaborate with others. And these points are not met I think.

As Wrangler I have good experience collaborating with Dan. He's reporting what he's he working on and progress of his Features/Changes, communicating what makes sense or not etc. so I agree with Rex - he made great progress.

I'd ask Wolfgang about his opinion on that matter, as someone who worked closely with Dan, his insight would be more useful than anyone's here.
https://bugzilla.redhat.com/show_bug.cgi?id=891282

Hmm, after everybody talk here in this ticket about me, i will clarify some things.
My relationship to Dan is very good since i step back in springtime to fedora for packaging MATE for fedora.
Much better than last year in summer, we could solve our difficults.

But this should be not the question?

Personaly i expected from a provent packager that he is able to build every package in fedora without any big issues.

These proposals were agreed upon on the yesterday's FESCo meeting:

  • AGREED: close ticket and ask if anyone sees misuse of powers, let us
    know (+8 if I count nirik's own implicit, -0, 0:0) (t8m, 18:15:29)
  • AGREED: change text to: "A FESCo member will send an e-mail to the
    sponsors list for the packager group to review the application.. You
    must get at least 3 positive votes with no negative votes, over a
    one week review period, to be automatically approved.. In case of
    negative votes, after that period FESCo will vote at one of its
    weekly meetings on your request and then notify you if your request
    has been approved." (+9, -0, 0) (t8m, 18:19:45)

Login to comment on this ticket.

Metadata