[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: trends.debian.net updated [request for additional groups/plots…inc. salvaged pkgs]

Hi Jonas and -devel,

Jonas Smedegaard <jonas@jones.dk> writes:

> Quoting Mattia Rizzolo (2020-04-11 17:20:48)
>> On Sat, Apr 11, 2020 at 11:10:48AM +0200, Wouter Verhelst wrote:
>> > On Fri, Apr 03, 2020 at 10:41:55PM +0200, Lucas Nussbaum wrote:
>> > > https://trends.debian.net/ was just updated (with data until April 1st).
>> > 
>> > There is a significant bump in the number of co-maintained packages
>> > during the buster release cycle. It is not at all clear to me what
>> > happened there.
>> > 
>> > Do you have any idea?
>> My assumption is that the deprecation of alioth lead many "team" formed
>> by 1 or 2 people to be replaced by simply comaintained packages.
>> That, together with the the @packages.d.o maintainer address, I reckon
>> those might also be considered "comaintained" instead of "team
>> maintained".
> The introduction of the Salsa "debian" area - with its accompanying 
> explicit rules of being more team-ish than the similar area in Alioth - 
> was indeed a game changer for many packages which I had previously 
> maintained in smaller teams.  I guess (and hope) that's the case for 
> others too.  Not sure how, but it might be possible to check that thoery 
> if possible to gather statistics of how many packages was in the Alioth 
> debian area compared to the Salsa debian area.

If possible, I would also be interested in seeing the following plots on
the Co-maintenance chart:

  1. team-maintained
  2. co-maintained Alioth collab-maint -> co-maintained salsa "debian"
     * shared plot
  3. not co-maintained collab-maint -> not co-maintained "debian"
     * shared plot
  4. co-maintained not in collab-maint nor "debian" group
     * shared plot
  5. not co-maintained and not in collab-maint nor "debian" group
That said, it seems like it might be potentially unkind and group-thinky
to say that bus_factor=1 AND not in the "debian" group is a
"smell"...since there are people who prefer to work alone...nonetheless,
I'm curious how many of these packages exist!

Possibly OT: it would also be cool to gather BTS stats for ITSs, and in
particular it would be cool to see the rate of salvaging packages, and
to see how many are going to teams, to "debian", and to somewhere else.
We're coming up on two years of the existence of salvaging, so it feels
like it might be a good time to do this.


Attachment: signature.asc
Description: PGP signature

Reply to: