I'm trying to get my head around today's Rust keynote mess (https://thephd.dev/i-am-no-longer-speaking-at-rustconf-2023).
I could be very, very wrong here, as I don't know volunteer org dynamics properly, but:
Why aren't names being named?
Who said what, on behalf of whom, in what role, is important to have transparency & accountability for power. (If you don't want your name on it, maybe it's not good?)
Obviously there's some over-simplication, as demonstrated by the TM fiasco...
Setting individuals up for a dogpile is bad, so giving ownership of contentious decisions to a committee seems reasonable - but by doing that you get to make sure it really is a committee decision, and the ownership remains clear.
OTOH, I don't get why the TM proposal was so badly communicated, but the most likely explanation for that it's hard & I'm missing something.
Still... there seems to be a transparency issue, and I can't tell how to solve it without identifying decision-makers.
I found fasterthanli.me's view interesting: https://gist.github.com/fasterthanlime/42da9378768aebef662dd26dddf04849
It's spurred a few thoughts in me:
1) I'm not pessimistic about the Rust org yet. A 100% SLO on human behaviour is unrealistic, what matters to me is the incident response and improvements.
2) Oh darn, a blameless incident review focusing on systems, not people, is hard to reconcile with the transparency of naming everyone involved.
In a normal IR at work, names can be used because of the assumption of good intent...
but that's not going to happen on an open internet discussion, even if everyone thought they were following process or doing the right thing.
(I'm assuming no-one was trying something they thought nefarious, because how dumb would you have to be to think no-one would notice, given all this?).
And... even in a "blameless" incident review, very occasionally there's just bad behaviour that needs addressing. (*)
(*) Concrete example, paraphrased:
"We are only allowed to do X in an emergency. I want to do X, I think it would be good, overall. Can I gently nudge the system into emergency so I can do X?"
"No. If you do that, deliberately finding a loophole to bypass policy you disagree with, 'blameless' will not save you."
They listened, it all turned out fine. Sometimes it's worth checking the odds on forgiveness even if you're not requesting permission.
Anyway, 3) I believe in back channels.
A lot of this mess seems to have come with the misuse of back channels.
As a manager, I spent a lot of time chatting with individuals or small groups about stuff involving a large group. It's efficient and effective.
The point, though, is that the public, wider discussion is genuine: If a decision is made with a wider group, it really is made there. They're not bypassed, it's no rubber stamp.
Back channels should only assist official channels.
> Back channels should only assist official channels.
This approach makes me think of (my layman's view of) court proceedings: there's a lot of various nonpublic channels, but all actually meaningful things have to happen in open court, or via prompty publicly readable documents.
