sphalerite changed the topic of #nixos-dev to: NixOS Development (#nixos for questions) | NixOS 19.03 released! https://discourse.nixos.org/t/nixos-19-03-release/2652 | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html https://r13y.com | 19.03 RMs: samueldr,sphalerite | https://logs.nix.samueldr.com/nixos-dev
WilliButz has quit [*.net *.split]
catern has quit [*.net *.split]
nwspk has quit [*.net *.split]
elvishjerricco has quit [*.net *.split]
dmj` has quit [*.net *.split]
hl has quit [*.net *.split]
tv has quit [*.net *.split]
hl has joined #nixos-dev
hl has joined #nixos-dev
hl has quit [Changing host]
elvishjerricco has joined #nixos-dev
simpson has joined #nixos-dev
nwspk has joined #nixos-dev
tv has joined #nixos-dev
nwspk has joined #nixos-dev
nwspk has quit [Changing host]
WilliButz has joined #nixos-dev
dmj` has joined #nixos-dev
catern has joined #nixos-dev
Synthetica has quit [Ping timeout: 252 seconds]
sorear has quit [Ping timeout: 258 seconds]
vdemeester has quit [Ping timeout: 252 seconds]
Synthetica has joined #nixos-dev
vdemeester has joined #nixos-dev
sorear has joined #nixos-dev
cjpbirkbeck has joined #nixos-dev
orivej has quit [Ping timeout: 248 seconds]
alp has quit [Ping timeout: 258 seconds]
justanotheruser has quit [Quit: WeeChat 2.4]
justanotheruser has joined #nixos-dev
alp has joined #nixos-dev
alp has quit [Ping timeout: 252 seconds]
layus has quit [Quit: ZNC 1.7.3 - https://znc.in]
andi- has quit [Remote host closed the connection]
layus has joined #nixos-dev
andi- has joined #nixos-dev
cransom has quit [Quit: WeeChat 2.3]
cjpbirkbeck has quit [Quit: Quitting now.]
Synthetica has quit [Quit: Connection closed for inactivity]
simpson has quit [Ping timeout: 250 seconds]
simpson has joined #nixos-dev
Jackneill has joined #nixos-dev
Jackneill has quit [Remote host closed the connection]
Jackneill has joined #nixos-dev
Jackneill has quit [Read error: Connection reset by peer]
Jackneill has joined #nixos-dev
alp has joined #nixos-dev
<arianvp> About the whole kernel LTS debate from yesterday...
<arianvp> why don't we piggyback on another distro's patches?
<arianvp> like. Ubuntu's support windows seem in sync with those of NixOS. why don't we allow for using Ubuntu's kernel?
<arianvp> wouldn't that solve the whole "How are we going to support a kernel for 6 months" issue?
<arianvp> I think that's a more realistic option than finding a company to invest money in doing our own kernel fork
<domenkozar[m]> flokli: it does, money goes to paypal/stripe while opencollective does the accounting
<domenkozar[m]> accounting for transparency purposes, not legal reasons.
<domenkozar[m]> so it's going to say things like
<domenkozar[m]> flokli donated 20$ to NixOS foundation
<domenkozar[m]> Graham reimbursed 1000 EUR for Hetzner gmbh
<domenkozar[m]> etc
Synthetica has joined #nixos-dev
justanotheruser has quit [Ping timeout: 258 seconds]
orivej has joined #nixos-dev
justanotheruser has joined #nixos-dev
marek_ has quit [Changing host]
marek_ has joined #nixos-dev
marek_ is now known as marek
<gchristensen> arianvp: easier still is to just support what upstream supports, until a company wants us to do that work
{^_^} has quit [Remote host closed the connection]
psyanticy has joined #nixos-dev
{^_^} has joined #nixos-dev
<gchristensen> samueldr: have you noticed a successful eval recently?
<gchristensen> there are no waiting jobs, which is sketchy
<domenkozar[m]> anything for weekly? :)
<gchristensen> I should finish a blog post before the next weekly :/
<gchristensen> I used to work on blog posts when flying, but arianvp has been good at supplying me with airplane-sized improvement requests
<etu> domenkozar[m]: But you have a bunch now? :)
<domenkozar[m]> still, would love to have more :)
<domenkozar[m]> rfc meeting and weekly, busy Nix day :)
<arianvp> gchristensen: my pleasure
<gchristensen> arianvp: btw did you see the --no-net option?
<arianvp> nope not yet
<arianvp> link?
<gchristensen> it is by eelco. I'll try to find one.
<arianvp> very nice
<arianvp> This will save me
<arianvp> Speaking of, we should really have a manpage for nix 2.0 :)
<gchristensen> +100
<arianvp> oh no. we have a arbitrary code execution CVE for VLC coming up :(
<gchristensen> woooo
orivej has quit [Ping timeout: 268 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 268 seconds]
<domenkozar[m]> would appreciate someone to proof-read :)
<Taneb> domenkozar[m]: tungsten.io in the last item should be a hyperlink
<Taneb> The job ad feels wrong to me but I can't put my finger on it
<gchristensen> domenkozar[m]: let me get you slightly more interesting text for Tungsten
<domenkozar[m]> m'kay
<domenkozar[m]> Taneb: well it continues from the title
<domenkozar[m]> if you mean the same thing that bothers me :)
<gchristensen> ```[Introducing nixpkgs-tungsten: The most convenient way to working with Tungsten Fabric](https://dev.cloudwatt.com/en/blog/introducing-nixpkgs-tungsten.html)
<gchristensen> [Tungsten Fabric](https://tungsten.io) is the open-source offering of [Contrail](https://www.juniper.net/us/en/products-services/sdn/contrail/) by Juniper Networks - a powerful SDN solution used by many big names in the IT industry. The [nixpkgs-tungsten](https://github.com/cloudwatt/nixpkgs-tungsten) project provides tools and workflows that make it much easier to work with, and on Tungsten Fabric itself.
<gchristensen> [Cloudwatt](https://www.cloudwatt.com/) is using it in production very successfully.```
<gchristensen> domenkozar[m]: ^
<domenkozar[m]> pushed
<domenkozar[m]> shipit?
<gchristensen> looking
<gchristensen> domenkozar[m]: shipit
<gchristensen> thank you!
<domenkozar[m]> ole :)
orivej has joined #nixos-dev
<samueldr> gchristensen: no, during yesterday's afternoon/evening I lazily checked and there were none, on maybe 6 tries
<samueldr> last successful eval of trunk-combined the 25th at ~23:26 UTC
pie_ has quit [Ping timeout: 258 seconds]
alp has quit [Ping timeout: 264 seconds]
psyanticy has quit [Quit: Connection closed for inactivity]
cransom has joined #nixos-dev
cransom has quit [Client Quit]
cransom has joined #nixos-dev
mingc has quit [Quit: WeeChat 2.4]
<samueldr> I'm at 4 or 5 today, still hasn't eval'd succesfully
<gchristensen> good grief
<gchristensen> are they al like nixos:trunk-combined, Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS ?
<samueldr> exactly
<samueldr> I would have notified on anything else
<samueldr> so, no OOM
<gchristensen> thanks
pie_ has joined #nixos-dev
Drakonis has joined #nixos-dev
<das_j> join #icinga
<das_j> ffs
<gchristensen> its a trap
<das_j> wha?
<gchristensen> icinga
<samueldr> :/ github being half-useless with PR checks being all wonky on a merged PR
<gchristensen> what's up?
<samueldr> #63156 might have broke eval, but am unsure if the checks passed
Jackneill has quit [Remote host closed the connection]
<{^_^}> https://github.com/NixOS/nixpkgs/pull/63156 (by Izorkin, 1 week ago, merged): phpfpm: do not run anything as root
<etu> samueldr: It did pass the checks
<samueldr> good, thanks
<etu> it was all green and stuff
<samueldr> no worries, here I'm just double-checking things, but the issue here is that github checks go all wonky once a PR is merged, so it's hard to be sure :)
<etu> :)
<etu> yeah, that's a bit annoying
<etu> samueldr: oh
<etu> samueldr: If you press "view details"
<etu> samueldr: On the merge message
<samueldr> yes, but see all the "Details" links?
<samueldr> they are added even though there was no details initially
<samueldr> so I don't have much trust in the results post-merge
<etu> uhh
<etu> yeah, dead links everywhere
<gchristensen> build logs?
<etu> no, just the links that github outputs links to the PR
* samueldr is waiting on the hydra-eval-jobs running locally
<samueldr> when there is a real link, it works, but where there was no link, it adds a dummy link
<etu> gchristensen: https://github.com/NixOS/nixpkgs/pull/63156 -- Click "View details" on my merge there, it all links back to the PR
<{^_^}> #63156 (by Izorkin, 1 week ago, merged): phpfpm: do not run anything as root
<gchristensen> wow
<Drakonis> that looks like a risky pr
<samueldr> btw, etu
<samueldr> error: undefined variable 'fpmCfg' at /nix/store/pnlg9bycirawpry9gx99yfnymszzlki5-source/nixos/modules/services/web-servers/phpfpm/default.nix:77:25
<samueldr> from trunk-combined
<etu> Then I missed something :/
<samueldr> default values
<samueldr> weird though how it didn't hig in the eval for trunk-combined.nix
<samueldr> didn't hit*
Synthetica has quit [Quit: Connection closed for inactivity]
<gchristensen> samueldr: you think it isn't memory pressure related?
<samueldr> no suspicious `with`, fpmCfg is not defined :)
<samueldr> though waiting on a non-memory-constrained eval from my workstation to double-confirm, just in case
<etu> hmm
<etu> I'm not sure what to put as default values there...
<samueldr> s/fpmCfg/cfg/ ?
<samueldr> though I haven't checked
<samueldr> ah, no, wouldn't make sense
<etu> That would make it recursive
<etu> I'd say that the package one should point at the default package
<samueldr> though, this is phpfpm.pools.phpPackage, not phpfpm.phpPAckages
<samueldr> (mkRemovedOptionModule [ "services" "phpfpm" "phpPackage" ] "Use services.phpfpm.pools.<name>.phpPackage instead.")
<samueldr> hm
<samueldr> so no, that's not it, it's a removed option
<etu> And lines should probably be emptystring then, so people can append their own settings
<etu> So I made #63867
<{^_^}> https://github.com/NixOS/nixpkgs/pull/63867 (by etu, 40 seconds ago, open): nixos/phpfpm: Remove usage of undefined variable fpmCfg
<etu> I think that should be fine
<das_j> gchristensen: It actually is a trap :(
<gchristensen> what happened?
<gchristensen> samueldr: where are you seeing that error? I see a different error
<samueldr> I don't have a repro locally yet :)
<gchristensen> I was going to try stopping the queue runner to see if that helps
<samueldr> well, the hard drive of that machine being full sure won't help /:
<samueldr> (my machine)
<pie_> any suggestions before I open a PR to add mapDir and mapDirFiles to lib?: https://bpaste.net/show/914596798856
<pie_> actually, brb, I have to reboot because my mount points are messed up
pie_ has quit [Remote host closed the connection]
pie_ has joined #nixos-dev
pie_ has quit [Remote host closed the connection]
pie_ has joined #nixos-dev
pie_ has quit [Remote host closed the connection]
pie_ has joined #nixos-dev
<gchristensen> samueldr: hydra-eval-jobs is up to 20.2g here lol
<samueldr> VIRT or RES?
<clever> gchristensen: ive had a nixops eval take 15-20gig before
<gchristensen> both
<samueldr> because here VIRT is ~32GiB, but RES doesn't increase much
<samueldr> gets killed once it goes above 1100M RES
<gchristensen> evaluator_initial_heap_size = 20000000000 evaluator_max_heap_size = 22000000000
<clever> gchristensen: at least interpolate the strings in nix!
<gchristensen> lol
<clever> gchristensen: evaluator_initial_heap_size = ${toString (20 * 1024 * 1024 * 1024)} is much more readable
<gchristensen> I was surprised it went right to 20g so fast, but I guess, duh, I told it to
<gchristensen> it worked samueldr
<samueldr> great
<gchristensen> I suppose it'd be flying too close to the sun to try turning on aarch64 :P
<gchristensen> ok I'm going to let the evaluator run for another 30min or so, and then revert the config change (it can't run this much ram in the evaluator, and the queue runner at the same time)
<gchristensen> so how about we put chef in to hydra as a build machine and fetch a new one with a few more gb of ram
<samueldr> get right in the sun please
<gchristensen> good grief hetzner prices are silly
<gchristensen> 128G RAM, 67eur/mo
<pie_> can someone review a small pr?: https://github.com/NixOS/nixpkgs/pull/63871
<{^_^}> #63871 (by deliciouslytyped, 33 seconds ago, open): lib: add mapDir and mapDirFiles
<gchristensen> oof
ma27_ has quit [Quit: WeeChat 2.4]
ma27 has joined #nixos-dev
Drakonis has quit [Ping timeout: 246 seconds]
Jackneill has joined #nixos-dev
<makefu> pie_: normally new library functions are only added if existing code is simplified with the new lib function
orivej has quit [Ping timeout: 268 seconds]
<pie_> welll, my use case is mapping import over a directory, which gets rid of files that are just a big list of imports
<pie_> I'd want it for https://github.com/deliciouslytyped/nix-rootedoverlay/tree/clean-up , which in theory *would* simplify a lot of code. im still working on it though.
<pie_> I don't have any existing code to point to but there's gotta be people that want a mapDirs :P
Drakonis has joined #nixos-dev
<gchristensen> that is not a pattern we'd want to encourage in Nixpkgs itself, so not sure it should live in Nixpkgs
<pie_> whats the alternative
<gchristensen> aye, that is a uqestion
<pie_> i dont see why this is necessarily bad
<gchristensen> ie: why we wouldn't want to do this in nixpkgs?
<pie_> yeah
Drakonis has quit [Ping timeout: 252 seconds]
<pie_> for stuff with plugins im creating a plugins subdirectory and using this instead of creating a file which is just full of imports
Jackneill has quit [Remote host closed the connection]
orivej has joined #nixos-dev
alp has joined #nixos-dev
<pie_> heh, here's a much older implementation of what I just did and its basically the same thing: http://chriswarbo.net/projects/nixos/useful_hacks.html
<pie_> coincidence? I think not :P
<pie_> * scroll down to "Importing Directories"
<gchristensen> yeah, it could be useful in some scenarios, but it isn't appropriate for Nixpkgs