<{^_^}>
#113848 (by benwbooth, 12 hours ago, open): nixos programs.sqlite missing large numbers of packages
mkaito has joined #nixos-dev
<siraben>
Is there anyway to get a breakpointHook equivalent on darwin?
<maralorn>
I am concerned about this commit, which broke stable for syncthing users. Breaking our stable channel with deprecations sounds like something that we should really avoid. https://github.com/NixOS/nixpkgs/commit/38eaa62f28384bc5f6c394e2a99bd6a4913fc71f (It's okay to mark go as vulnerable but not before fixing packages which depend on it.)
<sterni>
maralorn: it was discussed here already yesterday or something and I believe the consensus was to revert it
<sterni>
maybe we should actually revert the go_1_14 thing
BaughnLogBot has quit [Ping timeout: 260 seconds]
BaughnLogBot has joined #nixos-dev
mkaito has quit [Quit: WeeChat 3.0]
ekleog has quit [Quit: WeeChat 2.9]
ekleog has joined #nixos-dev
cole-h has joined #nixos-dev
AlwaysLivid has joined #nixos-dev
AlwaysLivid has quit [Remote host closed the connection]
AlwaysLivid has joined #nixos-dev
<worldofpeace>
Just want to remind everyone that there still isn't a new assigned release manager. And atm probably no candidates that I've been in contact with that I've approved of
<worldofpeace>
I would say we might be at the phase where we might need to bring in a rm from a previous release
tilpner_ has joined #nixos-dev
tilpner has quit [Ping timeout: 272 seconds]
tilpner_ is now known as tilpner
<worldofpeace>
and I for one am not in a place where I would be able to be involved with that process again (at least so soon). So if u can ask around. Jon did say if needed he could do it alone for 21.05, since he's managing really well with RFCs and everything else, but what's most important is that there is a successor within the timeline that we expect for there to be successors. If Jon is RM again he would have to do it once more to educate the
<worldofpeace>
person for the next release, or we're going to have to bring in both of us to do it simultaneously (a prior RM doesn't have fresh relevant knowledge anymore so they wouldn't be able to keep the current continuum as much). Plus that changes the schedule of when a RM's term ends. At the new year ur free from ur obligation since there's two releases in a year. So from then on the tradeoff happens in the middle of the year instead of the
<worldofpeace>
end.
<worldofpeace>
I know for sure that the impact of covid has a lot to do with this and I see a lot of people who could be candidates burnt out or disinterested in that kind of work
rajivr has quit [Quit: Connection closed for inactivity]
<worldofpeace>
In many ways I would say that people being disinterested in being an RM have valid points for not wanting to, but also someone has to herd the sheep and you can't be timid to do that
<worldofpeace>
And in my experience RM is not exactly the best word for what the project actually needs or what u end up doing. The project more so needs a "project leader", evaluating bugs was a tiny part of what I had to do, and more so I was delegating etc. like a leader for someone to recognize in nixos
<worldofpeace>
anyways, end rant need halp π€£
abathur has joined #nixos-dev
BaughnLogBot has quit [Ping timeout: 265 seconds]
BaughnLogBot has joined #nixos-dev
<domenkozar[m]>
worldofpeace: maybe we could ping top20 most frequent contributors in last release?
manveru has quit [Ping timeout: 272 seconds]
manveru has joined #nixos-dev
mkaito has joined #nixos-dev
<worldofpeace>
domenkozar: I did ping a couple of the frequent contributors
<worldofpeace>
but still, being a frequent contributor is not always equivalent to u being a good candidate as a release manager. I actively tried to be one so it would make it an obvious choice because I really wanted the position
<domenkozar[m]>
I agree with that, but if we don't have a good candidate we could lower our expectations :)
<colemickens>
I wonder if there would be more likely would-be RMs if there was also a semi-committed team of people ready to assist with ZHF/PR review.
<domenkozar[m]>
each release I see a lot of people helping out, but they are not really dedicated in the traditional sense
* colemickens
nods
<colemickens>
400 reviews in a month is a lot, on top of the leadership, wrangling, tracking, etc.
<colemickens>
I guess I'm just expressing that it's hard to feel like you're signing up for the latter without putting yourself in the line of fire for the former. No one's fault necessarily, just my perception.
<domenkozar[m]>
I wonder if paying for release management would help
<domenkozar[m]>
as in, the person doing the coordination would get compensated
<domenkozar[m]>
worldofpeace: did any of your candidates say why they wouldn't do it?
jonringer has joined #nixos-dev
srk has quit [Ping timeout: 268 seconds]
<worldofpeace>
domenkozar: the process, governance, CI, time commitment and just not being comfortable with role itself were all relevant
srk has joined #nixos-dev
<worldofpeace>
domenkozar: paying could help perhaps, tbh these companies use and want nixos people could pay it forward if they want to support the project
<worldofpeace>
overall I'm very realistic that it's not a perfect or optimal position and could be stressful and confusing
<domenkozar[m]>
hmm I'd love to dive deeper into those reasons to see what we can change
<worldofpeace>
and far as lowering expectations... like it's a position with a lot of expectations and needs improvement. one could easily be overwhelmed
<domenkozar[m]>
nothing is set in stone really, we can fix the the process and governance
<domenkozar[m]>
worldofpeace: should we have an audio chat about this?
<domenkozar[m]>
it feels to me it's not a topic for text :)
<worldofpeace>
<domenkozar[m] "hmm I'd love to dive deeper into"> I'm aware and I think Jon is also pretty aware and each release we've been addressing these issues. but it's the kind of thing that's difficult to do while you're trying to release simultaneously.
<domenkozar[m]>
yeah :/
<domenkozar[m]>
how about we schedule a call?
<worldofpeace>
We should, yeah
<worldofpeace>
(rn is not optimal since I'm with friends)
<domenkozar[m]>
would be great if Jon can also join
<worldofpeace>
he's in the channel π jonringer
<domenkozar[m]>
alright I'm available most of GMT/UTC working hours (or even late in the evening) so you let me know what time fits :)
<worldofpeace>
17:00 GMT domenkozar ?
<worldofpeace>
monday, lol forgot the actual date which is important
<domenkozar[m]>
that sounds good, jonringer ?
<worldofpeace>
I messaged him on discord because I don't think he really actively watches irc
<worldofpeace>
domenkozar: he said yeah
<worldofpeace>
I wonder if anyone else would want to be in on this, open to ideas clearly π
ris has quit [Read error: Connection reset by peer]
ris has joined #nixos-dev
<domenkozar[m]>
indeed :)
jonringer has quit [Remote host closed the connection]
<{^_^}>
rfcs#79 (by dasJ, 17 weeks ago, closed): [RFC 0079] No more direct pushes to master and release branches
<ashkitten>
domenkozar[m]: why did you do that in the first place anyway? i'm upset that you didn't talk to maintainers before pushing an update like that.
<domenkozar[m]>
breaking backward compatibility on minor releases is not something I'd expect from software
<domenkozar[m]>
ashkitten: but you're completely right
<domenkozar[m]>
it's a wrong assumption to have in the first place
<ashkitten>
it's api and abi compatible, only the configurations were changed
<lovesegfault>
The debate on "what constitutes a breaking change" is older than goat and I don't think we'll make any breakthroughs here :P
<domenkozar[m]>
isn't configuration an api?
<domenkozar[m]>
yeah I'd prefer to avoid that as well :)
<worldofpeace>
Valodim: *does a little curtsy* Much appreciated β¨ Difficult but insightful
<domenkozar[m]>
ashkitten: I'm curious what changes does configuration need?
<supersandro2000>
there was also a recent silent revert of a new test without any notice
<ashkitten>
there are more config files now, and i believe they removed the env var that the module uses to specify the config path
<supersandro2000>
most if not all changes should go through a PR to make sure that it at least evals
<domenkozar[m]>
I agree - no need to debate this :)