Re: (semi-)automatic unclaim of packages with more than 2 weeks of inactivity (and missing DLAs on www.do)

[Sylvain Beucler <beuc@beuc.net>]

Hi,

On 20/08/2020 19:37, Holger Levsen wrote:
>>> p.s.: as an after thought re: "don't harass me" (though I get it was a
>>> joke, but I think the joke conveyed a useful notion): maybe my semiautomatic
>>> mails should have a permanent disclaimer that being 'called out' by them is
>>> nothing bad and doesn't deserve any explaination, just fixing? I've just took 
>>> a note to do so next monday, please help me to word this disclaimer nicely.
>> Whether the e-mail is from a human or from a machine, whether it has a
>> disclaimer or not, it's never pleasing to be called out.
>> We better do it only when there's a valid reason.
>> Which I think is the case for the prior parts of the weekly e-mail.
> well, reserving an DLA and not sending it is also a valid reason.
>
> and I still disagree with being called out here is a bad thing, because it's
> worse to claim a package and not do the work. So being called out^w^wreminded
> of this is still a good thing in my book.
>
> we work in public here and this means our good and bad work is public.
> if one cannot work like this, one should learn to cope or switch.
>
> that said, I'll edit the results to not include missing DLAs on webwml if the
> DLA was requested on day I'll be sending the mail.

As a general guideline, I try to steer away from public shaming, which
is detrimental to team spirit and tends to hide real issues, and steer
towards problem identification and correction.

In this particular case of missing web imports, one real issue is a
fragile workflow involving duplicate mail/web announcements due to lack
of automation/integration.

Still in this particular case, in our process the team coordinator cites
contributors by running a heuristic-based script, and forwarding it
verbatim to the team (and the whole Internet), so I believe this isn't a
case where the contributor would need to learn to cope, but a case where
public naming is error-prone and not appropriate. Hence why I replaced
the last-committer name with package+date for missing website imports.

That being said, I think the current process for reporting stalled and
multiples claims is good enough as-is :)

Cheers!
Sylvain