<jtojnar>
e.g. I reviewed the code in this commit, I read the changelog associated with this change and did not find anything objectionable, I read the upstream commit log, I read these changed files in the upstream repository…
<samueldr>
without really knowing, my gut tells me that could be such an out-of-band method
<samueldr>
though really I'm not exactly sure what this will end up doing
<jtojnar>
interesting, looks like it is very wip at the moment
bennofs__ has joined #nixos-dev
<jtojnar>
I wonder if other distros use some structured tool to annotate and review upstream code and related documents
<gchristensen>
might be interesting to see, though I think most repos don't use a monorepo and use more delegation per package?
<qyliss>
lots of distros do
<qyliss>
void, alpine, Homebrew, and all the BSDs come to mind
<gchristensen>
great point
bennofs_ has quit [Ping timeout: 265 seconds]
<jtojnar>
I guess for GNOME updates, we could actually just paste the changelogs into etherpad and then cross out uninteresting stuff and highlight things that need work inline
<jtojnar>
but that would be tedious
<jtojnar>
so yeah, want some groupware that can take different data sources and allows to conveniently annotate them
<gchristensen>
qyliss: I wonder about social norms in those projects around maintainership
<qyliss>
well I used to be a Homebrew maintainer so I can talk about that at great length, but I don't really know about the others
<qyliss>
In Homebrew, there were maybe 10 maintainers, who were collectively responsible for both the package manager and the packages
<qyliss>
everybody had access to everything, although obviously some people focused more on one part or the other
<qyliss>
and then there were lots of non-maintainers (Homebrew uses "maintainer" to mean what we'd call "committter") who would send PRs
<qyliss>
so there was no real concept of maintainership of any specific part of the system
<qyliss>
everybody co-maintained all of it
<qyliss>
although there was a lead maintainer who was additionally responsible for infrastructure, dispute resolution, and overall vision
<gchristensen>
gotcha
<supersandro2000>
and reviewing?
<qyliss>
reviewing was done by the maintainers also
<qyliss>
at the time I was around, we had one very prolithic maintainer who did most of the PR reviews and package updates
<supersandro2000>
brew had less PRs at that time?
<qyliss>
it probably got 10-15 a day
<infinisil>
I feel like "vision" is a keyword that nixpkgs lacks
<qyliss>
from my experience with Homebrew, I think that that's a good thing.
<infinisil>
qyliss: To have one or to not have one?
<qyliss>
not having one
<infinisil>
Oh, why so?
<qyliss>
too much power in too few hands
<infinisil>
I don't think having a vision is tied to that
<qyliss>
I think it is
<gchristensen>
nixpkgs has 1,000 people with vision all executing at once
<gchristensen>
quite a sight to behold
<qyliss>
yes
<qyliss>
there's still vision, just no overall vision that can be enforced
<qyliss>
or at least not much of one
<supersandro2000>
I have the feeling vision always ends in a religious discussion
<jtojnar>
there are also areas where there is a vision (e.g. Nix) and they are heavily affected by the visionary’s bandwith
<qyliss>
Nix has too much vision for my liking
<jtojnar>
right. I would not mind that much if we could try to steer it but at the same time I do not want to steal Eelco's already pretty limited time
<supersandro2000>
nix the language, nix the program or nixpkgs? I am confused
<ekleog>
gchristensen: IMO we're missing a team-based structure to make vision be more in-sync
<ekleog>
I've started thinking about it on-and-off, and currently thinking that having a few teams that would each be allowed to have vision in their own domain and then people are not members of all the teams and must abide by the relevant team's vision would be best (as well as more closely match how Rust does things)
supersandro2000 has quit [Disconnected by services]
<ekleog>
Drawback being, team meetings take time, but this hopefully would be balanced by the gained coordination, as it's the natural price for syncing vision between multiple individuals
supersandro2000 has joined #nixos-dev
<ekleog>
So well, still something I'm thinking about and haven't prepared anything concrete yet, but if it can give people ideas… :)
<infinisil>
ekleog: As much as I know some people won't like this idea, I'm all for it. Feels kind of necessary for Nix to step up to the next level
{^_^} has quit [Remote host closed the connection]
{^_^} has joined #nixos-dev
contrapumpkin has joined #nixos-dev
copumpkin has quit [Killed (rothfuss.freenode.net (Nickname regained by services))]
contrapumpkin is now known as copumpkin
disasm has quit [Ping timeout: 276 seconds]
disasm has joined #nixos-dev
mkaito has quit [Quit: WeeChat 3.0]
rajivr has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
rajivr has quit [Ping timeout: 264 seconds]
grfn has quit [Ping timeout: 260 seconds]
raboof has quit [Ping timeout: 260 seconds]
feepo has quit [Write error: Connection reset by peer]
davidtwco has quit [Read error: Connection reset by peer]
thoughtpolice has quit [Read error: Connection reset by peer]
dmj` has quit [Read error: Connection reset by peer]
pmy_ has quit [Read error: Connection reset by peer]
rajivr has joined #nixos-dev
davidtwco has joined #nixos-dev
feepo has joined #nixos-dev
raboof has joined #nixos-dev
grfn has joined #nixos-dev
dmj` has joined #nixos-dev
thoughtpolice has joined #nixos-dev
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
justanotheruser has joined #nixos-dev
justan0theruser has quit [Ping timeout: 265 seconds]
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
orivej has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
jonringer has quit [Remote host closed the connection]
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
BaughnLogBot has quit [Ping timeout: 240 seconds]
BaughnLogBot has joined #nixos-dev
pmy_ has joined #nixos-dev
pmy_ has quit [Read error: Connection reset by peer]
pmy_ has joined #nixos-dev
orivej has quit [Ping timeout: 265 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 264 seconds]
saschagrunert has joined #nixos-dev
kini has quit [Remote host closed the connection]
kini has joined #nixos-dev
<eyJhb>
Is there any function in nixpkgs, that takes a input package e.g. php, and then only "get" the "bin" dir? Currently dockerTools will take all the dirs in e.g. PHP, and then it will copy them into the contaner. So this is the huge mess I am guetting - https://termbin.com/c7ysq
<adisbladis>
cole-h: I take it eyJhb's problem is the merging of other paths like etc, lib, share etc
cole-h has quit [Ping timeout: 256 seconds]
<eyJhb>
adisbladis: Yeah that is the issue, however I found that if I don't directly add stuff to contents, then it will not do that. Ie. if I create a init script, and add that, then it will have php, nginx, etc. in the /nix/store, but not pollute /
teto has joined #nixos-dev
<piegames>
supersandro2000: Never heard of graphql ^^ But the problem is not with borg, but when somebody else does the review
betawaffle has quit [Ping timeout: 240 seconds]
mdlayher has quit [Read error: Connection reset by peer]
mdlayher has joined #nixos-dev
davidtwco has quit [Read error: Connection reset by peer]
davidtwco has joined #nixos-dev
betawaffle has joined #nixos-dev
__monty__ has joined #nixos-dev
<roberth>
eyJhb: you probably want to use something like `buildEnv { paths = [pkgs.php]; }` to put in your docker image root
justanotheruser has quit [Ping timeout: 258 seconds]
mdlayher has quit [Read error: Connection reset by peer]
betawaffle has quit [Read error: Connection reset by peer]
davidtwco has quit [Write error: Connection reset by peer]
mdlayher has joined #nixos-dev
davidtwco has joined #nixos-dev
betawaffle has joined #nixos-dev
justanotheruser has joined #nixos-dev
ris has quit []
_ris has joined #nixos-dev
<asymmetric>
fyi i was having issues with 3.nixos.pool.ntp.org, not sure who's running those servers
<ajs124>
so. lua. are there any good reasons why we default to a release that's been EOL since 2015? or is it the usual case of nobody noticed or had the time to switch to a newer release?
<ajs124>
or is this maybe not a problem with lua because it's magical somehow?
<lassulus>
maybe everything breaks if you try to update it? :D
<ajs124>
I'd assume a bunch of stuff breaks, but having a modern default with things that need the old thing pinned to that seems how we normally go about this, instead of defaulting to the old thing.
<abathur>
arianvp: not sure if it's meaningful to contact the Apple dev-rel, but we've been getting reports of what looks to be an ongoing issue with nixbld users in Big Sur (nix#4531). I've got a PR in with a migration script and an installer fix going forward (nix#4532); some baked-in pain for daemon users who don't notice this before updating; can ref FB8997501
<{^_^}>
https://github.com/NixOS/nix/issues/4531 (by abathur, 1 day ago, open): macOS Big Sur updates w/ multi-user Nix present "booted to Recovery" because "an error occurred migrating user data"
<samueldr>
and without state it'd be hard(er) to generate those percents in the table I gather
<gchristensen>
yeah
<samueldr>
that's why I started with the dumb list suggestion
<samueldr>
knowing your love of state
<gchristensen>
hehe
<gchristensen>
I gave https://channels.nix.gsc.io/ a lovely facelift (lol) after moving it to S3 (which doesn't enumerate directories)
<samueldr>
yeah, good enough
<samueldr>
it's not like it's marketing copy
<gchristensen>
not yet!
<gchristensen>
thanks for the ideas!
<gchristensen>
maybe each report could generate an fragment, and then $(cat */fragment) in to a list
<samueldr>
that's a good option, stateful~ish though
<gchristensen>
S3 is basically a blockchain though, so it is okay
pmy_ has quit [Ping timeout: 272 seconds]
pmy_ has joined #nixos-dev
LnL- is now known as LnL
justanotheruser has quit [Ping timeout: 260 seconds]
_ris is now known as ris
<ris>
curl -H "Accept: application/vnd.github.v3+json" https://api.github.com/repos/NixOS/nixpkgs/releases -> empty array. turns out we don't publish releases, just auto-marked pseudo-releases based on tags
__monty__ has quit [Quit: leaving]
danderson has quit [Quit: leaving]
<gchristensen>
right, the releases aren't very useful anyway - they're essentially bookkeeping so `git describe` works