Re: [gentoo-project] Email from comrel -> Your recent contributions to the gentoo mailing lists

[Michał Górny]

On Wed, 7 Dec 2016 03:24:16 -0800
Daniel Campbell <zlg@g.o> wrote:

> On 12/07/2016 01:05 AM, Michał Górny wrote:
> > On Tue, 6 Dec 2016 23:24:22 -0800
> > Daniel Campbell <zlg@g.o> wrote:
> >
> >> On 12/06/2016 11:15 AM, Seemant Kulleen wrote:
> >>> The larger philosophical question is: Are we seriously banning people as
> >>> a community? What sort of community are we, in fact?
> >>
> >> I think that's largely speaking for itself here.
> >>
> >> It's clear that (a lack of) communication in this case and in Ian's did
> >> not prove to be fruitful. I hope Gentoo can improve in its community
> >> management in the future.
> >>
> >> I did not respond to William's e-mails mostly because they were
> >> off-topic in my thread, but I hold no ill will against him and feel he's
> >> just as entitled to a voice as anyone else. I think we're doing Gentoo a
> >> disservice (especially in Java land) by not having him help us. I
> >> disagree both with the decision to ban him (to outside readers: the
> >> greater developer community was not asked for their opinion) and the
> >> manner that William made his points (it was too much, too often).
> >>
> >> These happenings have severely dampened my view of the management within
> >> Gentoo and I will be voting accordingly next year.
> >
> > You should really have tried working with him before forming
> > an opinion. Let me put up a few facts for you. Some people may actually
> > end up seeing some similarity between the two incidents.
>
> To be fair, his exclusion from Gentoo kinda prevents me from having the
> opportunity to work with him -- at least on Gentoo-related things. That
> said, I'm open to any new information or facts presented. Before we get
> into nitty gritty I want to thank you for taking the time and
> consideration to reply so thoroughly and civilly.

Well, just to be clear, he is not forbidden from contributing using
the various technical matters available, if he only wished to
(and behaved at least a little).

> > First of all, William's technical competencies were, to say, lacking.
> > He may be very good in Java land but he is lacking in basic shell, not
> > to mention ebuild. His ebuilds end up doing the kind of three-flip
> > workaround for issues he himself introduced. If he was left to commit
> > freely, other people will end up having a lot of work cleaning up
> > after him.
> >
> > Of course, every deficiency can be solved if one wants to learn.
> > However, William more than once shown that he's stuck in some point
> > in the past and refuses to move forward. If you try to teach him, soon
> > enough he's going to change to subject to either how important he is
> > (and therefore he doesn't need to learn, you should fix stuff you
> > think he does wrong), how bad Gentoo is these days or plainly to
> > offending you. I think you have seen a fair sample of that on the ml
> > lately.
> >
> > So no, William is not suitable for commit access to ::gentoo. We really
> > don't have the manpower to fix all those issues, and I'm already tired
> > enough after one ex-developer leaving packages that could have never
> > worked (i.e. ebuilds that simply end up tripping on some 'die'
> > in the eclass that was always there) -- and we (the Python team) kept
> > hitting on those even a year later.
>
> Are we talking the sort of errors that a quick bug report and/or IRC
> ping should be able to fix, or a severe QA report that we all know we
> shouldn't do (like something covered in the quizzes?). Regardless, it
> seems fair to expect someone to be receptive (and hopefully learn
> something) when they screw up. I agree that we can't have people at
> Gentoo who aren't able to see flaws in their work and try to correct
> them. Creating work for other devs is counteractive to our goals.

I'm afraid we're talking about all kinds of errors unless they blow up
in his face. Including those sneaky silent errors that aren't directly
noticeable but can cause true indirect headache. Obvious security holes
like doing 'chmod 777' included (note: I haven't looked for any proof
of this, it was told to me).

> Would you or someone else be willing to show a few bug reports or
> examples of things he's done that caused us a headache? I'd search
> myself, but I don't really have much to go on.

As I've already given you on IRC, the best example would be:

https://github.com/gentoo/gentoo/pull/201

>
> > Finally, William tends to harass people who step on his toe, which can
> > happen pretty easily (i.e. via attempting to make him learn, see
> > above). If you do, you can expect to be privately and publicly harassed
> > for quite some time, at least to the point when he switches to
> > harassing someone else.
> >
> > Believe it, it's really not nice or productive to be highlighted every
> > 15 minutes by some irrelevant, jerky comment. Or be a topic of every
> > fourth mail. If you don't believe, you should try it.
>
> Oh, I fully understand that. I've not been on the business end of that
> particularly, but interruptions are a huge pain when you're trying to
> focus on anything relating to code. I hope that we agree on that one. :)
> >
> > As a side note, I should point out that we already had a similar case
> > in the past. I don't know if William would resort to that as well but
> > the other person even went as far as lying to other Gentoo users
> > and making them harass you privately.
> >
> > So yes, maybe Java lost some. Gentoo may have lost a few potential
> > developers too. However, I believe that Gentoo was saved from a major
> > loss of developers and contributors which has already happened once due
> > to the previous person mentioned, and I'm pretty sure would follow
> > William's staying longer with us.
> >
> > Hope this clears all the missing facts. I should point out that it's my
> > personal opinion, based on what I've seen and heard. Hope I didn't
> > betray anyone's confidentiality.
> >
>
> I think you've helped me see a bit more of the picture. I'm reluctant to
> turn anyone away; I trust our usual methods of contributing are still
> open to him (as they are for everyone else), should he have a change of
> mind or heart. My thoughts on management remain, and I think better
> information sharing (facts, evidence) can prevent huge threads like this
> from happening. If further evidence confirms William's behavior then
> I'll have to retract my statement regarding his contributing capacity.
>
> Thanks again for sharing with me.

Well, it would be really nice if he wanted to contribute using our
push-after-review method. However, so far he has been focusing on
requesting direct push access and pretty much a 'free hand' to do
whatever he wishes.

-- 
Best regards,
Michał Górny