> > The biggest questions for me are: Who makes decisions right now? Who
> > is handling money? What's the overlap? I know there's a desire for
> > collective decision making, which is great, but right now I think a
> > smaller group of core people (Ludovic + some others) needs to put a
> > structure in place because it feels like nothing will happen
> > otherwise. A little bit of benevolent dictatorish action could really
> > get the ball rolling here.
>
>
> Exactly. It's a really tricky situation. I think we all look to Ludovic
> when something needs to happen. And I don't think he (hi!) needs to feel
> that kind of pressure.
this is also a chicken-egg issue: until some authority is delegated, the center
will remain a bottleneck.
and this applies to a lot of things guix: from decision making to the monorepo
(which is holding an ever-growing flow of boring package update commits that is
crowding out guix infrastructure changes). or the single debbugs instance
washing together 20.000 pending package updates, and 15 interesting discussions
on how to improve guix itself (figurative numbers).
meanwhile there seems to be a growing inflow of enthusiastic and inspired users
who are bombarding the castle. this shows up in various forms, like the growing
patch review backlog, or the growing frustration expressed on the mailing list.
the bandwidth issue of the center won't be resolved without a way to delegate
compartmentalized authority to subteams (e.g. a python channel in a separate
git repo, and the gnome channel deciding which python channel commit to depend
on). and god forbid, maybe even allow them to chose their preferred git forge!
i'd keep the guix core to the minimum that can bootstrap a console-only system,
and then let a million channels bloom. some could be maintained or just blessed
by the core team, some may merely be listed, while others ignored.
unfortunately the infrastructure would need to evolve to accommodate this (e.g.
eliminate the dual registry of packages; consider promoting package definitions
from being a general toplevel form in a scheme file to something a little more
specific and a bit more constrained; IOW consider reifying the package
database, introduce package namespaces, syntax for package lookup, etc. maybe
reuse the guile module system for this, but in a less permissive way than it is
currently used). this obviously needs to be well thought out, but i don't even
see it considered, let alone mentioned as desirable.
a tangential to illustrate the above:
at this point i was wondering where could be the list of ideas/vision for the
future of guix, to see whether such a thing was ever considered. then i found
the TODO file. then i saw that it's very outdated and rather untidy. then i had
an impulse to add some guix-devel archive links to the distributed substitute
discussions, and also to delete or mark some entires DONE that have long been
implemented. then i considered the effort it would take to send a patch, and
the fact that i already have a lot of my effort bitrotting away in the issue
tracker, and why would this one not also fall through the cracks... and then i
decided not to. and that untidy TODO file will remain there to be one of the
inputs that newcomers will use to form an impression, which will then inform
their decisions about e.g. whether to contribute.
if it were a wiki page, or a file in a separate guix-doc repo, etc, then giving
me write access would not be the same decision as giving me commit access to
the guix repo.
to sum it up: human cooperation cannot grow beyond a certain size without an
organizational structure that can accommodate the newcomers. my impression is
that guix has reached such a threshold. and while a new stucture is not formed,
potential contributors are constantly frustrated away. and it's very much not
obvious to judge how much value is lost that way.
--
• attila lendvai
• PGP: 963F 5D5F 45C7 DFCD 0A39
--
“One of the evils of paper money is that it turns the whole country into stock
jobbers. The precariousness of its value and the uncertainty of its fate
continually operate, night and day, to produce this destructive effect. Having
no real value in itself it depends for support upon accident, caprice, and
party; and as it is the interest of some to depreciate and of others to raise
its value, there is a continual invention going on that destroys the morals of
the country.”
— Thomas Paine (1737–1809), 'Complete Writings of Thomas Paine' (1786)
