Re: RFR: email about regressions [was: Dealing with ci.d.n for package regressions]
Emilio Pozuelo Monfort writes ("Re: RFR: email about regressions [was: Dealing with ci.d.n for package regressions]"):
> On 25/05/18 12:24, Mattia Rizzolo wrote:
> > But rather than false positive I should have said "things a maintainer
> > can usually do ~nothing about", like RC bugs on other package where NMUs
> > are infaseable, or the migration is blocked by some transition, etc (in
> > those case it's not like the information is false, it's simply not
> > useful).
>
> Even if there's not much one can do in some cases, I think it's still useful to
> know that your package is scheduled for autoremoval.
Yes, I agree, and on the whole I want the autoremoval mails so much
that I have grep-excuses emailing me daily.
It is irritating when then RC bug was in another package, and has been
fixed there, and the autoremover and the testing migrater are simply
racing (and the latter is going to win).
> > Anyway, doesn't really change my point that the autopkgtest mails are
> > welcome to me :)
>
> Certainly.
Ultimately we have to make a guess here as to what is better for most
maintainers. We might treat the package with tests (the delayer) and
the package being migrated (the delayee) differently.
As regards the delayer, ISTM that most of the relevant maintainers
have opted into autopkgtest by providing test cases and are likely to
want to know when they fail. I doubt that it is these maintainers who
will complain.
As regards the delayee, it is less obvious. These people haven't
opted in. One could say that justice demands that they be told the
reason why their package is being delayed. But of course if they
care to look in the excuses, the information is there already.
My argument as regards the delayee is rather different: delayee
maintainers may be annoyed at the impact on their work. In such
situations it is normally a good idea to communicate early, rather
than late; and an automatic mail is a good way to set some behavioural
expectations.
Ian.
--
Ian Jackson <ijackson@chiark.greenend.org.uk> These opinions are my own.
If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.
Reply to: