Re: [gentoo-project] The problem of defunct and undermanned projects in Gentoo

[Daniel Campbell]

On 07/16/2017 02:12 PM, Michał Górny wrote:
> Hi, everyone.
>
> I'd like to re-heat a worrying topic. More than once I've noticed that
> some projects in Gentoo are either completely defunct or seriously
> undermanned. This has been already tackled via the herd->project
> conversion but the problem still exist and I'd like to try to solve it
> better.
>
> According to a quick grep, we have 175 teams defined right now. With
> a quick grep, I get the following histogram:
>
>   Members   Teams
>         0       3
>         1      36
>         2      34
>         3      19
>         4      20
>         5      22
>         6      12
>   ...
>
> Besides the first two numbers, it doesn't look that bad. However, most
> of the time the number of members does not map well into real activity.
>
> I think some of the projects could really do with just new members.
> However, not having active developers on the project makes it hard for
> someone to actually issue a call for members.
>
> I'm thinking of mailing every project in Gentoo (i.e. mass-mailing 175
> projects, some developers will get a lot of mail). The mail would
> contain some descriptive text and a list of project members. It would
> serve two goals:
>
> 1. asking the project to verify its member list (some people still don't
> keep their wiki up-to-date),
>
> 2. asking the project if they need me to send a call for members.
>
> Projects that are entirely fine could reply with a simple 'ACK'. Others
> could ask for help directly. Those who don't reply will be brought to
> further consideration.
>
> What do you think? Would you mind getting that amount of mail once?
> Any other ideas?
> Getting some new members -- specifically community members who know more
about a given project's topic -- would be very beneficial to Gentoo,
imo. I think some of our projects suffer from Gentoo devs who want to
keep projects (or rather, their packages) alive but lack expertise in
that particular domain, so difficult or odd bugs take longer to diagnose
and fix than they would if there was a project member who was familiar
with the software. The alternative is of course disbanding the
project(s) and dropping the packages, but to me that should be used as a
last resort.

Which of our outreach avenues has proven the most effective? If we have
a call for contributors, are we equipped to onboard enough people --
proxy-maint or otherwise -- to bring activity to these smaller projects?

I'm admittedly not a people person, so I'm not sure how I can help.
-- 
Daniel Campbell - Gentoo Developer
OpenPGP Key: 0x1EA055D6 @ hkp://keys.gnupg.net
fpr: AE03 9064 AE00 053C 270C  1DE4 6F7A 9091 1EA0 55D6