worldofpeace_ changed the topic of #nixos-dev to: #nixos-dev NixOS Development (#nixos for questions) | NixOS stable: 20.03 ✨ https://discourse.nixos.org/t/nixos-20-03-release/6785 | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html | https://r13y.com | 19.09 RMs: disasm, sphalerite; 20.03: worldofpeace, disasm | https://logs.nix.samueldr.com/nixos-dev
ris has quit [Ping timeout: 272 seconds]
gleber has quit [Ping timeout: 240 seconds]
gleber has joined #nixos-dev
alp has quit [Remote host closed the connection]
rajivr has joined #nixos-dev
puck has quit [Ping timeout: 264 seconds]
puck has joined #nixos-dev
bgamari has quit [Ping timeout: 246 seconds]
bgamari has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 265 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-dev
justanotheruser has quit [Ping timeout: 258 seconds]
jonringer has joined #nixos-dev
jonringer has quit [Ping timeout: 264 seconds]
vdemeester has quit [Read error: Connection reset by peer]
vdemeester has joined #nixos-dev
dongcarl has joined #nixos-dev
jared-w_ has quit [Ping timeout: 258 seconds]
jared-w_ has joined #nixos-dev
<samueldr> jtojnar: sorry to single you out, but you're the font wizard and stained that way forever
<samueldr> on a 20.03 system, I can run nix-shell -I nixpkgs=channel:nixos-unstable -p piper --run piper, quit, then run something else and the font cache is invalid
<samueldr> leading to things like quassel segfaulting
<samueldr> Fontconfig warning: FcPattern object weight does not accept value [40 210)
<samueldr> luckily simply `rm -rf .cache/fontconfig` (or fc-cache -r) fixes that
<samueldr> but to me this looks like... a pretty bad problem
<samueldr> not only quassel btw, pretty much anything using fonts in any sort either hard crashes or uses squares
<samueldr> (about to look at github if there are existing issues, otherwise open one)
orivej has quit [Ping timeout: 264 seconds]
<samueldr> #97418 about the problem ^
<{^_^}> https://github.com/NixOS/nixpkgs/issues/97418 (by samueldr, 4 minutes ago, open): Running unstable apps on 20.03 can corrupt the font cache
* samueldr prepares for further tests
<samueldr> and checked that it's new from the latest unstable bump, and it is
<samueldr> anyone that updated to the latest bump of unstable can test running a 20.03 program?
<cole-h> samueldr: I can try (except it'll be -small, not nixos-unstable)
<samueldr> well, as long as it's from after today's bump
<cole-h> Last updated 2 hours ago, should be good
<cole-h> Oof, 2.5G download :D
<cole-h> samueldr: https://github.com/NixOS/nixpkgs/pull/95358 is the most likely culprit
<{^_^}> #95358 (by jtojnar, 3 weeks ago, merged): fontconfig: Stop using versioned config dirs
<samueldr> I had that in the back of my mind, simply because of the name, but didn't want to cloud anyone's judgement by making half-assed suppositions :)
<cole-h> Oh, crap. Time to rebuild emacs-pgtk...
<cole-h> jk, just gonna disable that for now.......
<cole-h> Oh fork, I have to rebuild the kernel.
<cole-h> Maybe it's time to see how the rt patchset compares to zen
<colemickens> cole-h: that's in nixpkgs-wayland now btw
<cole-h> colemickens: emacs-pgtk?!
<colemickens> emacs-pgtk
<cole-h> colemickens++ +++++++++++++
<{^_^}> colemickens's karma got increased to 36
<colemickens> bqv did it, I just merged it
<colemickens> real credit to builds.sr.ht for building it tho, it's the true champ
<cole-h> True
<cole-h> sr.ht is fantabulous
jared-w_ is now known as jared-w
<etu> colemickens: That would be nice to mention in #nixos-emacs as well for people there, now I have to check out nixpkgs-wayland for that :D
<cole-h> Also, I think delft / eris (idk which) needs to be redeployed so that 20.09 shows on s.no.o?
<cole-h> samueldr: confirmed
<cole-h> Running 20.03 software as of the recent nixos-unstable breaks as well
<cole-h> (reverting, fc-cache -r'ing, and trying again, to verify)
<cole-h> Oh wait, I was on -small before, so yeah, still happens lol
<cole-h> Wait
<cole-h> samueldr: jk, my last unstable-small doesn't have the suspect fontconfig stuff in it
<cole-h> 0a64ff8152b46d691a0a087598a489c49752d142
<cole-h> idek anymore :D
cole-h has quit [Ping timeout: 258 seconds]
sascha has joined #nixos-dev
sascha is now known as Guest35326
peelz has quit [Read error: Connection reset by peer]
peelz has joined #nixos-dev
alp has joined #nixos-dev
jonringer has joined #nixos-dev
zimbatm has quit [Read error: Connection reset by peer]
zimbatm has joined #nixos-dev
zimbatm has joined #nixos-dev
zimbatm has quit [Changing host]
alp has quit [Read error: Connection reset by peer]
alp has joined #nixos-dev
jonringer has quit [Ping timeout: 256 seconds]
teto has joined #nixos-dev
peelz has quit [Ping timeout: 260 seconds]
__monty__ has joined #nixos-dev
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 265 seconds]
orivej_ has joined #nixos-dev
alp has quit [Remote host closed the connection]
alp has joined #nixos-dev
FRidh has joined #nixos-dev
orivej_ has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-dev
teto has quit [Quit: WeeChat 2.9]
<domenkozar[m]> evening
orivej has quit [Ping timeout: 258 seconds]
teto has joined #nixos-dev
justanotheruser has joined #nixos-dev
FRidh has quit [Quit: Konversation terminated!]
<worldofpeace> okay, so I'll have the ZHF issue open in "my evening" and also the discourse post
<aterius> There's also an emacs pgtk nativecomp branch
teto has quit [Ping timeout: 264 seconds]
teto has joined #nixos-dev
Guest35326 has quit [Remote host closed the connection]
evanjs has quit [Quit: ZNC 1.8.1 - https://znc.in]
evanjs has joined #nixos-dev
urkk_ has joined #nixos-dev
urkk has quit [Ping timeout: 265 seconds]
jonringer has joined #nixos-dev
teto has quit [Ping timeout: 265 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 246 seconds]
<nbp_> zimbatm: I am not sure about the value of having json/toml modules. However, I see it being valuable as a way to load from existing config files, for options values.
<nbp_> zimbatm: one nit, add a test case, there is a directory under lib.
<nbp_> I guess one advantage of having json/toml input for modules, is that there is plenty of tools to manipulate these file formats!
nbp_ is now known as nbp
alp has quit [Ping timeout: 246 seconds]
alp has joined #nixos-dev
<Ericson2314> I am wary of baking in module system in Nix for the same reason i am wary of flake
<Ericson2314> * I am wary of baking in module system in Nix for the same reason i am wary of flakes
cole-h has joined #nixos-dev
evanjs has quit [Quit: ZNC 1.8.1 - https://znc.in]
evanjs has joined #nixos-dev
<Valodim> heh. you're working on builtin ipfs support, but are wary of flakes?
<nbp> Ericson2314: I concur with these concerns!
<Ericson2314> Valodim: yes lol
<Ericson2314> there is a very careful layering argument behind the IPFS stuff
<Ericson2314> for example, it's not necessary that Nix need know about IPFS, but it is necessary that it know about the way IPFS hashes things
<nbp> Ericson2314: has the anonimity aspect of the requester/provider been investigated for IPFS integration?
<Ericson2314> nbp that is purposely out of scope
evanjs has quit [Quit: ZNC 1.8.1 - https://znc.in]
<Ericson2314> I can integrate things well but I cannot perform crypto miracles.
<nbp> ok.
evanjs has joined #nixos-dev
<cole-h> aminechikhaoui: I think we also need a redeploy of wherever status.nixos.org is hosted (still shows 19.03) -- delft / eris I think?
alp has quit [Ping timeout: 272 seconds]
alp has joined #nixos-dev
orivej has joined #nixos-dev
<aminechikhaoui> cole-h hm might be the case, let me check
alp has quit [Remote host closed the connection]
alp has joined #nixos-dev
ris has joined #nixos-dev
<aminechikhaoui> hm not sure why eris deployment triggers a Nix rebuild, maybe some flake magic
<cole-h> aminechikhaoui: Looks like it's because delft's nix input follows hydra's nix input, which is just short of master (e.g. probably not cached)
urkk_ has quit [Ping timeout: 240 seconds]
urkk has joined #nixos-dev
<aminechikhaoui> ok 20.09-small shows up now https://status.nixos.org/
<aminechikhaoui> I guess the others didn't have the "tested" job succeed yet
<cole-h> Weird -- 19.09 doesn't show as EOL, and 19.03 is still there
alp has quit [Remote host closed the connection]
alp has joined #nixos-dev
<cole-h> Ah, there wego
<cole-h> No more 19.03
<cole-h> (Though 19.09 not EOL'd yet)
<cole-h> Thanks aminechikhaoui -- I'm sure it'll get sorted over time
<aminechikhaoui> I did nothing to fix it :p
<cole-h> You redeployed it, didn't you? ;)
<aminechikhaoui> right, I was just talking about the EOL channels, I guess the service that does the update was still running
teto has joined #nixos-dev
sascha has joined #nixos-dev
rajivr has quit [Quit: Connection closed for inactivity]
sascha is now known as Guest94709
AlwaysLivid has joined #nixos-dev
justanotheruser has quit [Ping timeout: 258 seconds]
peelz has joined #nixos-dev
justanotheruser has joined #nixos-dev
Guest94709 has quit [Remote host closed the connection]
peelz has quit [Read error: Connection reset by peer]
peelz has joined #nixos-dev
alp has quit [Ping timeout: 244 seconds]
alp has joined #nixos-dev
__monty__ has quit [Quit: leaving]
page has quit [Ping timeout: 246 seconds]
kalbasit_ has joined #nixos-dev
justanotheruser has quit [Read error: Connection reset by peer]
page has joined #nixos-dev
disasm has quit [Ping timeout: 240 seconds]
disasm has joined #nixos-dev
justanotheruser has joined #nixos-dev
<worldofpeace> Okay everyone ZHF started https://github.com/NixOS/nixpkgs/issues/97479
<{^_^}> #97479 (by jonringer, 53 minutes ago, open): ZERO Hydra Failures 20.09
<infinisil> \o/
alp has quit [Ping timeout: 258 seconds]
<immae> Is it really necessary that `curl https://hydra.nixos.org/build/126812502/reproduce` downloads the whole nixpkgs git repository each time? (I’m trying to contribute for once, and I’m trying to find my way in how hydra works, but downloading 2GB of data for each test is a bit too much for my connection :/ )
<samueldr> those "reproduce" links are kind of an artifact of old times
<immae> ah
<samueldr> you can probably go 98% of the way by checking the inputs of the build
<immae> how should I run the tests then?
<samueldr> and just building that
<samueldr> so I assume it's https://hydra.nixos.org/build/126812502, right?
<immae> yes
<samueldr> you can know the revision of nixpkgs by checking https://hydra.nixos.org/build/126812502#tabs-buildinputs
<immae> ok
<immae> (nb: in the current case ejabberd builds fine, I suspect the test lacks some directory)
<samueldr> and for the configuration, top menu, jobset -> overview, then the Configuration tab https://hydra.nixos.org/jobset/nixos/release-20.09#tabs-configuration
<worldofpeace> hmm, why are we using the reproduce thingy? was it recommended somehow?
<immae> worldofpeace: I don’t know, I arrived on hydra and I try to find my way :)
<samueldr> I don't know that it's recommended, I simply assumed immae saw the link existed :)
<immae> clicking on "reproduce locally" seemed the best guess at that time :p
<samueldr> immae: see, there's "Nix expression:" in the Configuration tab?
<samueldr> that's the thing you can nix-build
<samueldr> nix-build to get the whole attrset of the jobset
<samueldr> so back on the build, the build is "nixos.____", well tests.ejabberd.x86_64-linux
<samueldr> is the attrname
<samueldr> aaaand... somehow I'm wrong
* samueldr checks
<immae> :D
<immae> (you lost me I’m re-reading the steps one by one)
<samueldr> oh
<samueldr> it's nixos.tests.ejabberd.x86_64-linux the attrname, duh
<samueldr> $ nix-build ./nixos/release-combined.nix -A nixos.tests.ejabberd.x86_64-linux
<samueldr> or, $ nix-build $JOBSET_EXPRESSION -A $JOBNAME
<immae> Ok I got the idea, let’s try that
<samueldr> stableBranch and supportedSystems are not really used by anything that matters for reproducing
<immae> does it matter that I’m on the same revision of nixpkgs as https://hydra.nixos.org/build/126812502#tabs-buildinputs ? I assume if it fails there then it fails on master too no?
<samueldr> kind of, and kind of not in your case :)
<immae> (unless someone fixed it inbetween of course)
<samueldr> exactly
<samueldr> it matters to *reproduce* the build exactly, but it's likely as you said that it's still an issue on master
<immae> ok I see
<immae> So you gave me all the information to reproduce the same errors, but I might not need to be exactly equal
<samueldr> right
<immae> I’ll note that, thanks!
<samueldr> it's also a good idea to check on master too since... well, someone could have fixed it!
<samueldr> you don't want to spend minutes, or tens of minutes fixing something that's already fixed!
<immae> right :D
AlwaysLivid has quit [Ping timeout: 246 seconds]
<immae> I’ll first try to fix this one (even if it is fixed already) because I "know" ejabberd and it will let me get used to the mecanisms
<immae> (and I never wrote any test in nixos so far so it might be a better way to start)
<immae> oh one question
<immae> where did you get the -A in nix-build ./nixos/release-combined.nix -A nixos.tests.ejabberd.x86_64-linux ?
<immae> Ah in the title
<samueldr> exactly, the job name is the attribute name
<samueldr> ah, no need to go to the jobset page to know which expression was used https://hydra.nixos.org/build/126812502#tabs-details
<samueldr> it's in the Details tab
<immae> Hmm
<immae> I don’t see it there :D
<samueldr> Nix expression:file nixos/release-combined.nix in input nixpkgs
<immae> ah this part
<samueldr> yeah, the attr name is the job name, there's no "need" to have it listed elsewhere
<samueldr> now, what's up with the reproduce-build scripts?
<samueldr> I think it might be useful for complex setups within hydras, and for non-nixos hydras
<samueldr> as it literally tries to execute the build as closely as to what hydra did
<immae> (note tht the reproduce-build script didn’t work :p )
<samueldr> but most of the time you can get by 98% of the way by reading a bit in the UI
<immae> yes a nix-build is fine actually :)