<gchristensen>
I feel liek I will have to pay for my sins of persuading cole-h to learn some perl
<jonringer>
perl was :( because I had to delete volth from last years post
<hexa->
awhh, meh
<siraben>
Is there a fork nixpkgs-review extension that also posts build logs?
<siraben>
Maybe supersandro2000 knows
<siraben>
a fork of*
XgF has quit [Ping timeout: 276 seconds]
marek has quit [Ping timeout: 246 seconds]
XgF has joined #nixos-dev
marek has joined #nixos-dev
tomberek has quit [Quit: Connection closed]
teto has quit [Ping timeout: 260 seconds]
tomberek has joined #nixos-dev
rajivr has joined #nixos-dev
justanotheruser has quit [Ping timeout: 276 seconds]
justanotheruser has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
cole-h has joined #nixos-dev
<ryantm>
I think this is the way for forward for converting the NixOS manual to Commonmark. Please review and support me if you think so too https://github.com/NixOS/nixpkgs/pull/107444
<{^_^}>
#107444 (by ryantm, 19 weeks ago, open): nixos/doc: add md-to-db.sh, convert "Building Your Own NixOS CD" to CommonMark
<samueldr>
ryantm: maybe we can wait up to at least right after the branch-off?
<samueldr>
making sure we have the maximum amount of time in unstable before it goes to a stable branch
<samueldr>
(since it's close by anyway)
<ryantm>
samu,,:eldr jonringer was asking about getting the release notes in Markdown. Not sure if he meant this release or next.
<samueldr>
let's have their opinion then :)
<samueldr>
but I don't think it'd be nice if it ends up creating a lot of work to fix issues in the stable branch down the line (not that I see any cause for concern)
<ryantm>
Makes sense to me
<samueldr>
I haven't reviewed the PR, but I'm generally in favor of actions towards tackling the migration
<samueldr>
at some point the generic brand adhesive bandage needs to be ripped off
<ryantm>
The only problem I see is pages converted from markdown to Docbook will potentially lose some fidelity so it might be difficult. But those sections could stay as Docbook if really necessary until the end.
<ryantm>
And I guess flokli s argument for doing it as one long running PR. I don't like long running PRs.
<rmcgibbo[m]>
i honestly have no idea why it doesn't pick up home-assistant. i've tried to figure that out before, and come up short.
<hexa->
ah meh
<rmcgibbo[m]>
basically `nix log {drvPath}` is returning nothing, i think. it's not truncation at the level of my code.
<hexa->
weird
<hexa->
just built home-assistant, and `nix log $out` works
<rmcgibbo[m]>
😂 strange.
<rmcgibbo[m]>
let me file an issue to track this. it's annoying
<hexa->
thank you!
<hexa->
drvpath as well
<rmcgibbo[m]>
maybe there's a race condition between the build finishing and `nix log` responding with the log... unlikely. i really don't know (i just looked through the code again, and i don't see anything).
cole-h has joined #nixos-dev
<hexa->
fwiw i have seen the same Problem with nixpkgs-review
<maralorn>
Does someone now about LC_ALL stuff in nix-shell and can maybe chime in on the discussion in this issue? https://github.com/NixOS/nixpkgs/pull/122023 I don‘t feel competent to judge what the correct solution here is. Should we just hard code the locale in the script in question?
<infinisil>
endocrimes: Should this maybe go to staging instead since it's quite a bunch of rebuilds?
<infinisil>
Or is master intentional to get the security fix in asap
<endocrimes>
go usually goes straight to master afaik
<endocrimes>
But yeah bouncing CVEs via staging seems not ideal
<samueldr>
I think it all depends on the severity (which I haven't looked at at all)
<samueldr>
well, some kind of cross product of severity × rebuild amount
<qyliss>
yeah
<qyliss>
at least for non-security fixes, go bumps should probably go via staging
<samueldr>
total of 4727 store paths changed across the three systems ofborg checks
<samueldr>
4727 go things are probably a drop in the bucket
<samueldr>
aren't go builds quick?
orivej has joined #nixos-dev
<qyliss>
depends ime
<samueldr>
at least 4727 *total* across the uh... four system... isn't that bad for a rebuild
<samueldr>
(forgot i686-linux)
<qyliss>
yeah, esp. for security impact
<endocrimes>
Go builds are usually p quick
<endocrimes>
for nix anyway
Jackneill has quit [Ping timeout: 268 seconds]
<maxine>
Go has gone straight to master for every single bump I've done, with this exact discussion repeating... 😅
Jackneill has joined #nixos-dev
<adisbladis>
My 2¢: ~2 years back bumping Go was a 2-3 hour rebuild on a decently powerful machine (epyc 7401)
<samueldr>
adisbladis: rebuild of every paths, or go itself?
<adisbladis>
samueldr: Every path
<samueldr>
that sounds like a trivial rebuild amount then
<adisbladis>
I'm sure we have a lot more packages now and it probably takes longer, but given the security impact I think it's a no brainer
<endocrimes>
It’s like... 4-6 hours on a 24t ryzen CPU usually, so that checks out
MichaelRaskin has joined #nixos-dev
midchildan has quit [Ping timeout: 260 seconds]
raboof has quit [Ping timeout: 260 seconds]
steveeJ has quit [Ping timeout: 260 seconds]
davidtwco has quit [Ping timeout: 245 seconds]
thoughtpolice has joined #nixos-dev
midchildan has joined #nixos-dev
davidtwco has joined #nixos-dev
steveeJ has joined #nixos-dev
raboof has joined #nixos-dev
dmj` has joined #nixos-dev
rajivr has quit [Quit: Connection closed for inactivity]
<rmcgibbo[m]>
I wish there was a nix-community OSX build box. Trying to debug ZHF darwin issues on a crappy 2016 macbook i pulled out of my closet is painful.
<sterni>
can also confirm that darwin builds don't get started for me
<sterni>
but I've seen them working recently so
tomberek has quit [Quit: Connection closed]
<rmcgibbo[m]>
thanks for the pointer https://github.com/jtojnar/nixpkgs-fastmac to run nixpkgs-review on a darwin zhf pr. seems to work? might be a good tool for lots of people.
<rmcgibbo[m]>
Jan Tojnar++
<samueldr>
rmcgibbo[m]: the matrix bridge breaks nick expansions :(
<samueldr>
jtojnar++ on behalf
<{^_^}>
jtojnar's karma got increased to 73
justanotheruser has quit [Ping timeout: 260 seconds]