[kde-community] Have repo maintainers opt-in for github mirroring (was: Re: Official KDE mirror on github)

Teo Mrnjavac teo at kde.org
Sat Sep 19 15:13:17 UTC 2015


On Saturday, September 19, 2015 16:34:59 Albert Astals Cid wrote:
> El Dissabte, 19 de setembre de 2015, a les 16:23:26, Teo Mrnjavac va 
escriure:
> > On Saturday, September 19, 2015 13:04:55 David Edmundson wrote:
> > > > > I was under the impression they were disabled by the options we had
> > > > > selected. Unfortunately that is not the case.
> > > > 
> > > > Thanks for clarifying on this.
> > > > 
> > > > I hope they can still be disabled.
> > > > 
> > > > They can't. I had spent some time looking before. Sorry.
> > > 
> > > However, we have solid hard data that it's a non-issue.
> > > 
> > > Gnome has been mirrored on github for nearly 2 years, in that time GTK
> > > has
> > > had a grand total of 4 pull requests over time.
> > > Most others (gedit, cheese, epiphany) have had 0.
> > > 
> > > Interestingly they have had literally hundreds of github "forks", which
> > > implies it has led to sustantiable numbers of patches back using the
> > > traditional methods
> > > 
> > > I've made a wiki page, which says how to turn a pull request into a
> > > reviewboard submission.
> > > https://techbase.kde.org/Development/GithubMirror
> > > 
> > > If we get any questions we can then just copy and paste that, and don't
> > > need to spend any time explaining. Bam, done.
> > 
> > Thank you David, for your get-things-done approach in this controversial
> > and tense situation. It is really much easier to solve than it seems from
> > all these threads.
> > 
> > I'm personally in favor of letting projects decide whether to allow GitHub
> > pull requests or not, but regardless of the final decision it is good to
> > already have practical solutions like this techbase entry.
> > 
> > I find it unfortunate that some long time KDE contributors feel that KDE
> > goals are threatened by all this. I understand their concerns, but I
> > assign
> > those concerns a different priority score. In fact, the inflexible policy
> > towards 3rd party (including proprietary) infrastructure and processes we
> > have in KDE deters me from bringing some of my own (currently
> > GitHub-hosted) work under the KDE umbrella, as this would hinder some very
> > productive working relationships with our downstreams and potentially
> > result in *less* free open source software being produced, deployed and
> > used.
> 
> That's something you have convinced yourself about, you don't have proof.
> 

This is not a repeatable experiment, of course I don't have science level 
proof. I even pointed out that my opinion stems from assigning different 
priority scores to the same concerns others have pointed out, concerns that I 
do recognize.

Definitive proof could only be through hindsight after proceeding. The fact 
that you complain about the lack of proof is a bit baffling, as proof of the 
outcome for this kind of hypothetical scenario is something that no one can 
possibly produce without a crystal ball. This is a social, people-herding 
issue.

The best I can give you is a maintainer's assessment, and that's what I wrote, 
based on the knowledge I have of said downstream relationships. You may agree 
or disagree, but I don't think it makes much sense to complain about lack of 
proof.

Cheers,
-- 
Teo Mrnjavac
http://teom.org | teo at kde.org



More information about the kde-community mailing list