maria-discuss team mailing list archive
-
maria-discuss team
-
Mailing list archive
-
Message #05175
Re: community contribution measurement page
On Sun, 22 Jul 2018 13:29:07 +0200
Sergei Golubchik <serg@xxxxxxxxxxx> wrote:
> Hi, Daniel!
>
> Wow, this is very informative. Thanks a lot!
You're welcome.
> It's nice to see how "First PR response" time got improved (in 2015?)
> after Svoj started handling PRs.
True. Was that a foundation goal at the time?
> Recent increase in PRs can, perhaps, be partially attributed to
> Tempesta and Codership?
Maybe, if i get the Team/Organisations going this will be a bit more
obvious.
> And, perhaps, partially because we're getting
> more PRs now and need to change internal procedures to be able to
> cope with them?
In addition to growth, there is a 95th percentile of 25 days for
open prs in contrast with the median of 160 days shows a number aren't
being reviewed/rejected/merged.
If you can come up with some chart/goal that aids the internal
procedures then I can either try to develop or publish it.
Possibly something around the time in a github PR review state, or
releases passed for an approved PR (even though these aren't currently
scraped).
> I don't understand the "PRs open and closed this month" chart, number
> of "opened PRs" seems to be always less than the number of "closed
> PRs", how is this possible?
Its a stacked graphs but you're right, something's odd there, I'll take
a look.
https://github.com/MeasureOSS/Measure/blob/master/widgets/repo/43_prsStackChart.js
> We seem to be accepting about 60% of all PRs? It looks like since 2016
> we accept a bit more than a half. I think it's a pretty good ratio,
> meaning we must have rather smart contributors who create really good
> pull requests :)
Given the amount of feedback and contributor behaviour to create a new
pr or git push, or merged/rebased in background (as Vicențiu
suggested), I think 60% is something to be proud of.
As a measure subject to varying effects, its relative performance
probably would need to be very careful about inferences made.
> On Jul 22, Daniel Black wrote:
> >
> >
> > Folks,
> >
> > I've prepared a community contribution measurement page:
> >
> > http://measure-mariadb.ozlabs.org/repo/mariadb/server-outside-org.html
> >
> > This is largely the code base of
> > https://github.com/MeasureOSS/Measure
> >
> > known errors:
> > * I've yet to create the organisation aspects of this site.
> > * some top page may be distorted by non-mariadb respositories
> >
> > Insights and suggested improvements welcome (especially with
> > code) :-)
> Regards,
> Sergei
> Chief Architect MariaDB
> and security@xxxxxxxxxxx
>
References