I have a few thoughts about this -- bearing in mind that I am a drive-by
contributor to Core, at best, and don't have much personal opinion other
than maybe "I wish it were easier to get stuff in".

1. I think that Antoine is correct that "it's easier and more natural"
is a bigger motivation for "office work" than is the fear of brigade.
So one thing is that any change to public processes shouldn't make it
_harder_ for people to collaborate online, since that could push
people more to in-person fora and we'd just have the worst of both
worlds. Or at least, anyone making such a change should have a lot of
confidence that the increased friendliness to earnest contributors
would outweigh the extra friction.

2. On the other hand, fear of brigades _does_ clearly have a nonzero
chilling effect. I certainly think about it when publicly communicating
near the project, and I commonly bring it up when doing things in
rust-bitcoin (i.e. "fortunately, we're not Core, so we can just do
[some change that would constrain wallet workflows, or which could
make ordinals particularly hard, or particularly easy, or whatever]"
and not have to worry about fallout.)

So at the very least, it's a factor that discourages some external
developers from being bigger contributors to the project.

3. And of course, it's not just obvious brigades -- when one or two
nontechnical people show up with strong political views about
something which really is not a political change (or at least,
doesn't have the political effect they believe it does, because of
their own misunderstanding), it's still discouraging and sometimes
stressful. And this happens all the time around mempool policy,
even if PRs with 100+ comments that get locked are fairly rare.

4. However, after (ironically) discussing this email off-list with a
bunch of people, I think that these problems stem from a fairly small
cultural issue: that the Github repo appears to be a totally open
forum where anyone is welcome to participate, even in code review
threads, because technically anybody _can_ participate with no
obvious sense that they're leaving X and entering somebody's
workplace.

And _this_, IMHO, might be solvable by something extremely simple. It
might be sufficient to just move from Github to Gitlab or Codeberg or
something where far fewer people have accounts. It would probably be
sufficient to just find a platform where you have to register on the
Core repo somehow then wait 24 hours before you can post, with the
implication that if you're not there to contribute technically, you
might lose your access. (This is true on Github but the only
mechanism is that you can be banned from the org, something that
feels heavy and scary for maintainers to use -- I really hate doing
this to non-bots on rust-bitcoin and I don't even have to worry that
they'll go on twitter to scream censorship and that I'm taking over
Bitcoin or whatever -- and is also more-or-less invisible to users
until it happens to them, so it's not an effective deterrent.)

It would certainly be effective to put a strong technical barrier,
e.g. you have to produce a custom mining share to join, or a strong
social barrier, e.g. you need personal invitations from two people.

But I think such tech barriers would be unnecessary and the social
barriers wouldn't be worth the cries of censorship and centralization
that they'd inevitably (and somewhat reasonably) cause.

5. I don't see much of benefit to making the repo *unreadable* to
outsiders. It sorta prevents linking on Twitter but if we expect
there to be mirrors, people can just link to the mirrors.


Again, it's not my project and I don't mean to advocate for anything in
particular. Just trying to organize thinking on the topic a bit.


--
Andrew Poelstra
Director, Blockstream Research
Email: apoelstra at wpsoftware.net
Web: https://www.wpsoftware.net/andrew

The sun is always shining in space
-Justin Lewis-Webster