[18581] in athena10
Re: Moving to salsa.debian.org/debian
daemon@ATHENA.MIT.EDU (Alex Chernyakhovsky)
Tue Sep 9 09:02:33 2025
MIME-Version: 1.0
In-Reply-To: <tslo6rjeqpl.fsf@suchdamage.org>
From: "Alex Chernyakhovsky" <achernya@mit.edu>
Date: Tue, 9 Sep 2025 09:02:17 -0400
Message-ID: <CAB18ysoe71r==fpy90vwPwV2P74H6TR+z8mWk4EJR9GErZKAVg@mail.gmail.com>
To: Sam Hartman <hartmans@debian.org>
Cc: Andreas Tille <tille@debian.org>,
Salvaging Team <team+salvage@tracker.debian.org>, debathena@mit.edu,
1098197@bugs.debian.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8bit
I'm not sure if I should be replying with my MIT/Debathena hat or my
Debian hat or my "inherited Zephyr upstream team maintenance after
Karl passed away" hat ... so I guess this message is all 3, and I've
chosen my MIT From: address arbitrarily.
After Karl passed away, Zephyr was transitioned to a Github "team" at
https://github.com/zephyr-im/zephyr. I'm one of the folks in that
organization, so we're upstream. This is a git repository that
superseded Karl's old subversion repository, with a conversion that
Karl himself performed, IIRC.
Karl had been working on a Zephyr 3.1 release before he passed.
Unfortunately, there was some sort of memory corruption bug we
couldn't trace, that only appeared on a "heavily used" zephyrd such as
the MIT zephyr ATHENA.MIT.EDU realm. I still have on my TODO list to
one day get around to fixing it, but no amount of ASAN builds showed
anything wrong.
There is a debian/ directory in this repository, but as Sam identified
in earlier messages, I do believe it has diverged from what's in
Debian upstream. I don't recall how much, but I suppose it is fixable,
from past memory.
To that end, if we need a new Maintainer, I (as achernya@debian.org)
am happy to do it -- although having it be team-maintained with the
other DDs and DMs who are knowledgeable would be better. Zephyr
traffic has reduced in the recent years, but MIT does still run
Zephyr, and does still use the upstream packages for clients, so I
think we should include it for now.
As to GitHub vs Salsa: I recognize that Salsa is the preferred Debian
way to do it, but we'd really need to stop carrying a debian/
directory in upstream if we want to go to Salsa. Karl being upstream
and also a DD led to it being included, and of course, later led to
the divergence.
Sincerely,
-Alex
On Tue, Sep 9, 2025 at 8:50 AM Sam Hartman <hartmans@debian.org> wrote:
>
> >>>>> "Andreas" == Andreas Tille <tille@debian.org> writes:
>
> Andreas> Hi Sam, I'd offer to fix RC bugs and do a QA upload while
> Andreas> the packaging can be maintained on Salsa in Debian/ team.
>
> As the closest thing to a maintainer, I will not stand in your way; if
> you want to do that, go for it with no delays.
> I'd ask you to consider whether that's really an efficient use of your
> time though if we're just going to drop Zephyr.
>
> Zephyr has been effectively unmaintained upstream since Karl's death,
> and I think that while the servers probably still run at MIT and a few
> other places, Zephyr traffic has died off significantly.
>
> Unless I'm wrong and someone steps forward and says they want to invest
> time, your effort might be better spent removing Zephyr as a dependency.
