At 2026-01-26T15:55:52+0100, Simon Josefsson wrote:
> "G. Branden Robinson" <[email protected]> writes:
> >> How did Debian end up like this?
> >
> > That's an _excellent_ question.  I propose that Debian Developers
> > (and Debian Maintainers) spend time thinking for themselves about
> > it.
> 
> There are many things involved here, but the social consequences of
> having a 'Maintainer:' field for packages, and no central version
> control system for development, explains a lot to me.

I don't agree.  While I can see how it's tempting to imagine that in
early years, Debian developers adapted the "maintainership"
concept to all administrative functions, that doesn't explain how we
ended up, even before the Debian Constitution was drafted, with "teams"
that weren't really "teams", but instead individuals who viewed
themselves as as functionally isolated from their teammates as they were
from any other Debian Developer.

Having now stated it thus, maybe you're right!

> Group maintainance is great, but it is a workaround for something
> else.

I once read, on the Web, someone crediting me with innovating the
concept of team maintainership when I announced the "X Strike Force" for
managing the unwieldy monolith of XFree86 packaging for Debian.[1]

I don't know if that's true.  I don't remember copying the idea from
anyone, but it could have been in the air.  What I recall of my
intentions was that the XFree86 package bug list was gigantic and after
some long duration (measurable in years), I felt I was making little
headway in burning that list down to zero.  I fixed many, many bugs, but
new ones seemed to come up with as much or even greater frequency.[2]

That was the reason for the name "X Strike Force".  Being a typical
American, I imagined that the solution to a persistent problem was to
bomb it into submission by bringing more resources to bear on it.

(Had I been a clever Brit, maybe I'd have taken a Churchillian approach
of employing chemical weapons rather than uncouth explosives.[4])

In any case, I'd say the primary and best justification for team
maintenance is the same as for any coöperative effort in labor: the
scale of the task exceeds the resources of a single diligent human.

Regards,
Branden

[1] https://lists.debian.org/debian-devel-announce/2003/06/msg00000.html

[2] The foregoing URL touts the adoption of Subversion for management of
    the package.  For my first few years maintaining XFree86, I did so
    without the benefit of any revision control system at all.[3]  As
    I've said before on this point--I was _completely insane_.  That, or
    CVS was so horrible to use that a nullity was preferable.

[3] A more sober claim might be that I used my Debian package releases
    as a snapshotting mechanism, and composed extensive change log
    entries to make a sort of audit trail.  At the time, some people
    grumbled about the length of those change logs just as they complain
    about my emails today.  Yo, people don't join up to READ STUFF.

[4] 
https://www.theguardian.com/world/shortcuts/2013/sep/01/winston-churchill-shocking-use-chemical-weapons

Attachment: signature.asc
Description: PGP signature

Reply via email to