From: Ariadne Conill <ariadne@dereferenced.org>
To: ~alpine/devel@lists.alpinelinux.org
Subject: Re: mailing lists, processes, modernization
Date: Wed, 22 Apr 2020 08:51:08 -0600
Message-ID: <1758765.52O7J0OIYB@localhost>
In-Reply-To: <66694f73048e047cadbe9e3f715801d8@dereferenced.org>
References: <66694f73048e047cadbe9e3f715801d8@dereferenced.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 7Bit
Content-Type: text/plain; charset="us-ascii"

Hello,

I would like to thank everyone for contributing their thoughts and general
discussion to this latest round of discussions about the current situation
with our mailing lists.

It seems that we have mostly gained consensus for using GitLab to track
planning-related development discussions.  Now, we will need to figure
out how best to accomplish this.  I appreciate everyone's patience while
we work through ideas and come to a consensus on how the GitLab-based
development planning will work.  There are a few possibilities that have
been discussed in IRC and elsewhere already, that I will outline in another
thread.

As you are likely aware, the initial position of this round was to propose
using GitLab for planning (which allows all sorts of advantages, such as
assigning deadlines and milestones to system change proposals), and to
move all other discussion to Discourse.

Well, feedback on using Discourse instead of a traditional mailing list
solution was quite mixed, but most contributors that contributed to
this discussion appreciated the goal of having a system that allowed
users to post via either e-mail or the web.

Which leads us to an alternative proposal for the mailing lists: adopting
mailman3 and hyperkitty.  Hyperkitty is a frontend to Mailman3 which
provides many of the same features we were looking for in Discourse.
If you're interested about the web experience, the fedora mailing lists
system uses Hyperkitty for user-facing components, which you can see
at https://lists.fedoraproject.org/.

In this case, we will still need to write a dumper for Sourcehut to convert
our current archives to mbox format so they can be imported into
Hyperkitty.  We will also need to migrate the subscriptions, obviously,
but both of these tasks should not prove that difficult.

Please let us know your thoughts on Discourse verses Hyperkitty as
a path forward for the remaining discussion that would not be well
served by using GitLab.

Thanks everyone,
Ariadne