Contributors mailing list archives


Re: Better contribution recognition

Holger Brunn
- 05/04/2018 09:44:41
> Personally for me, as far as I'd go is anyone who reviews a PR, raises a

> bug report, signs a CLA, has a PR merged, runs a repo, sits on the board

> etc etc in a given year gets an [Valued] OCA Contributor 2018 laptop

> sticker.  If Holger turns up with 12 stickers in 2030, people will know.

well, in line with my decay proposal every year somebody should go and
steal my sticker from the n-2nd year. But given a sticker only means
kudos, no need to let them decay. I like the stickers a lot.

> In terms of what I interpret Holger as saying, I don't want contributors

> who do nothing but raise PR's and never do any of the other work, or

> when they do they just give +1 reviews.  A leaderboard for contributors

> will only make that worse.  People already game app store numbers. 

> Making life easier for proven contributors (commit rights, lower review

> requirements, priority PR's as examples) by recognising the trust you

> have in them I think is far better.

Total agreement here. We nerds have the tendency to try and solve social
problems with technical means, which often is not the optimal solution.
But to be honest for big vastly anonymous communities with no fixed
boundaries as open source communities are, I have little ideas how to do
it differently.

I specifically like the priority PR proposal. No idea how to implement
this in github, that's why I proposed the blocking in case of too little
karma, that's simple to implement. But yes, as far as nudging goes, I
prefer rewarding wanted behavior over penalizing unwanted behavior most
of the time, here too. We just need to find a way to implement that.

But still, also here I think we need some more objective measure than a
popularity contest for determining who a proven contributor is. It does
already now feel to me that a rather small group (which includes me)
pingpongs too many PRs between each other, not even looking at the other
PRs. From my point of view, that's a rational way to deal with the
sparsity of my resources (time here), when I review a person's work of
whom I know she does good work, I'm sure I won't waste my time. With
other contributors, I don't know that, and given I've very little time
to spare on pure OCA work currently, I'm very risk averse here.
But globally that's horrible of course, because it closes off the
community to some degree and also gives room to favoritism in reviews.

> I mean, we need to put serious ourselves and try to professionalize this

> putting smart people in charge paying them to be in charge 5x7x7 at

> least just opening, finishing, merging, thinking about side effects and

> controlling the current chaotic stage of the work on the OCA.

I do have some slippery slope arguments here, but I see the point.
Especially looking through big repositories, sometimes also very
unpleasant discussions with contributors disagreeing with your decisions
as a maintainer, is quite serious work that I actually wouldn't want to
do for money more often than a few hours per month.

This could be a last resort option for repos like web. You can't do this
properly next to your daytime job. But also here, we first need a fair
assessment/measurement of whom we trust to be a good person for this
position. I don't think a new hire who has never worked with Odoo or
contributed to the OCA is a good candidate here.

Therp - Maatwerk in open ontwikkeling

Holger Brunn - Ontwerp en implementatie

phone: +31 (0)20 3093096