<NixOS_GitHub>
[nixpkgs] grahamc pushed 1 new commit to release-16.09: https://git.io/v1sX3
<NixOS_GitHub>
nixpkgs/release-16.09 60b89d7 Graham Christensen: guile: enable tests
glines has joined #nixos
ebzzry has quit [(Ping timeout: 248 seconds)]
stepcut has quit [(Remote host closed the connection)]
msd has joined #nixos
msd has quit [(Client Quit)]
marusich has joined #nixos
mguentner has quit [(Quit: WeeChat 1.6)]
PragCyph1 has joined #nixos
stepcut has joined #nixos
PragCyphr has quit [(Ping timeout: 260 seconds)]
stepcut has quit [(Remote host closed the connection)]
mguentner has joined #nixos
stepcut has joined #nixos
mguentner2 has joined #nixos
mguentner has quit [(Ping timeout: 248 seconds)]
ebzzry has joined #nixos
systemfault has quit [(Read error: Connection reset by peer)]
systemfault has joined #nixos
justanotheruser is now known as hairyengineer
hairyengineer is now known as justanotheruser
<eacameron>
Anyone tried to run their server behind cloudflare DNS using the built-in LetsEncrypt support? CloudFlare DNS messes my site up with the ACME module running.
<simpson>
eacameron: You're supposed to let CF do your TLS.
edvorg has joined #nixos
<simpson>
Yet another reason why I dislike CF, TBH.
<eacameron>
simpson: Yah that stinks.
<_habnabit>
you can pay a ton of money for something a bit more secure
<eacameron>
simpson: I'm not up on my TLS certificate stuff but I'm guessing CloudFlare's TLS is pretty low quality compared to a cert on your site.
<eacameron>
_habnabit: Like?
<simpson>
eacameron: AIUI they tick all the right boxes.
<_habnabit>
i mean CF's key server thing
<simpson>
I just dislike the centralization and all that.
<_habnabit>
it's not about quality; CF will score well on everything
<_habnabit>
they do PFS and all
<_habnabit>
but you opt-in to being MitM
<eacameron>
simpson: So there's really no security advantage to having the ACME module over CF...?
<eacameron>
I am a noob at this stuuf.
alx741 has quit [(Quit: alx741)]
<simpson>
eacameron: The advantage in my mind to not using CF is that you don't have to trust CF when not using them.
<eacameron>
simpson: LOL
<eacameron>
_habnabit: I see...
<simpson>
eacameron: No, I wasn't joking; if you use CF, then you necessarily trust them to not listen in on your connections.
<eacameron>
simpson: Yikes...I guess I didn't realize that. That royally sucks.
<eacameron>
I thought they just did DNS and caching...
<gchristensen>
they do
<eacameron>
And TLS was optional.
<gchristensen>
but necessarily they have to proxy and decrypt / encrypt / understand all your traffic
<_habnabit>
if you're only using CF for DNS, you don't need their TLS
<_habnabit>
but if you want caching or DDoS protection, ..
<eacameron>
_habnabit: All I care about is DDoS protection
<eacameron>
I know how to cache...
<_habnabit>
then you have to trust them with all your traffic
<eacameron>
gchristensen: I see. It's making sense to me now.
<eacameron>
_habnabit: Are there other ways to get DDoS protection?
<_habnabit>
use GCE/AWS
<gchristensen>
erm
<eacameron>
_habnabit: oh...I already use GCE
<gchristensen>
IIRC, AWS's DDoS protection philosophy was "scale up"
<eacameron>
LOL...
<_habnabit>
gchristensen, ELBs?
<gchristensen>
ELBs don't protect against DDoS
<gchristensen>
they let you scale up
<gchristensen>
there may be an updated whitepaper on it, but the last whitepaper from AWS on DDoS mitigation was simply scaling up
<jerzy>
how to install stratego now, (it seems not to be any longer part of all-packages) i am looking for pp-aterm (part of it) to pretty print derviation files
rly has joined #nixos
vbmithr has quit [(Ping timeout: 260 seconds)]
FRidh has joined #nixos
vbmithr has joined #nixos
Gohla has quit [(Ping timeout: 260 seconds)]
Gohla has joined #nixos
acertain has quit [(Ping timeout: 244 seconds)]
jgertm_ has quit [(Ping timeout: 250 seconds)]
DavidEGr1yson has joined #nixos
DavidEGrayson has quit [(Ping timeout: 260 seconds)]
nmattia has quit [(Quit: No Ping reply in 180 seconds.)]
aj__ has quit [(Ping timeout: 260 seconds)]
acertainkind has quit [(Ping timeout: 244 seconds)]
aj__ has joined #nixos
Guest66666 has quit [(Ping timeout: 260 seconds)]
nmattia has joined #nixos
acertainkind has joined #nixos
<NixOS_GitHub>
[nixpkgs] FRidh pushed 2 new commits to master: https://git.io/v1sAN
<NixOS_GitHub>
nixpkgs/master e39584c Frederik Rietdijk: pythonPackages.pyjwt: 1.4.0 -> 1.4.2
<NixOS_GitHub>
nixpkgs/master d224328 Frederik Rietdijk: pythonPackages.oauthlib: 0.7.2 -> 2.0.0
pstn has quit [(Remote host closed the connection)]
sdothum has joined #nixos
ebzzry has quit [(Ping timeout: 256 seconds)]
pstn has joined #nixos
johann__ has quit [(Ping timeout: 246 seconds)]
<NixOS_GitHub>
[nixpkgs] edolstra closed pull request #20834: rssh: Make rssh a valid shell (master...rssh) https://git.io/v1GKl
herzmeister has quit [(Quit: Leaving)]
herzmeister has joined #nixos
<gchristensen>
niksnut: there isn't a go-faster button, is there? :) one problem with this set of updates, is batching that critical firefox fix with world rebuilds means it takes a long time to get out the firefox fix.
<NixOS_GitHub>
[nixpkgs] FRidh force-pushed python-wip from 56db616 to 7038b0c: https://git.io/vzaOS
<NixOS_GitHub>
nixpkgs/python-wip 7038b0c Frederik Rietdijk: pythonPackages: do not propagate pytest...
<gchristensen>
I'm not sure splitting them would have made it more likely to get out sooner or not,though.
<viric>
Dezgeg: clever: are your pi systems booting without logind/journald problems related to SECCOMP? My system cannot start those daemons
<viric>
this actually menas that 1) I don't have system log and 2) logins take 20seconds
<Dezgeg>
master works for me
eacamero_ has quit [(Ping timeout: 244 seconds)]
pi3r has joined #nixos
<viric>
16.09
JonReed has joined #nixos
<viric>
Dezgeg: but you never had SECCOMP problems, yet I had to add SECCOMP to the kernel config
<viric>
why was it?
<viric>
I do have seccomp in the kernel... what else can be wrong?
<Dezgeg>
so CONFIG_SECCOMP_FILTER is y?
<viric>
It is a different error than the lack of SECCOMP; when SECCOMP was lacking in the kernel, I could not login
<viric>
Dezgeg: hm not in /proc/config.gz
<JonReed>
Hi, is there a way to see previous configurations.nix? I can rollback, but now I want to diff current configuration with the some old one, but did not `git init` at that time yet
<JonReed>
I can boot to it, but I can't find that configuration.nix actually is
<JonReed>
if it exists at all
<viric>
Dezgeg: damn it, I have OABI_COMPAT in the system.
<viric>
I messed my local git branches I guess
<viric>
interesting - I remember I could not login, before.
<Dezgeg>
yeah, I thought it showed up like that as well
<JonReed>
when it starts it shows "Default" "Configuration 1" "Configuration "41" "Configuration 45" etc. I want to see which configuration.nix produced "Configuration 41"
christine_ has quit [(Ping timeout: 245 seconds)]
FRidh has quit [(Read error: Connection reset by peer)]
FRidh has joined #nixos
johann__1 has quit [(Quit: Leaving.)]
jgertm is now known as Guest5872
Guest5872 has quit [(Killed (card.freenode.net (Nickname regained by services)))]
Mercuria1Alchemi has quit [(Ping timeout: 250 seconds)]
magnetophon has joined #nixos
Itkovian has joined #nixos
<NixOS_GitHub>
[nixpkgs] FRidh pushed 2 new commits to python-wip: https://git.io/v1Zvy
<NixOS_GitHub>
nixpkgs/python-wip c2130c6 Frederik Rietdijk: pythonPackages.pynzb: disable tests
<NixOS_GitHub>
nixpkgs/python-wip 7a49b43 Frederik Rietdijk: pythonPackages.webassets: disable tests...
erasmas has joined #nixos
<gchristensen>
has anyone read `man gitworkflows`? the part on a cherry-picking based workflow (similar to what we have) about merging stable into unstable regularly seems interesting
Itkovian has quit [(Read error: Connection reset by peer)]
<Unode>
currently trying without using a nix expression but simply nix-shell. Yet I can't get past the configure step. It fails to find lfs and posix. The first should be installed (lua52Packages.luafilesystem) but isn't found
<Unode>
this is also true if I run lua and try to "require('lfs')"
<Unode>
is there some additional setup needed to make this work?
<Sonarpulse-Work>
we need to replace magic or with ??
pstn has joined #nixos
<Sonarpulse-Work>
hard-to-reason about contextual keywords are the worst
<Sonarpulse-Work>
and then use ?? for default arguments for consistency
bennofs1 has quit [(Ping timeout: 258 seconds)]
<NixOS_GitHub>
[nixpkgs] danbst opened pull request #20839: module system: better option pretty-printing (add quotes when necessary) (master...fix-show-option) https://git.io/v1ZIo
rly has quit [(Ping timeout: 260 seconds)]
<shlevy>
Sonarpulse-Work: Patches welcome :) I doubt niksnut will want that though
<Sonarpulse-Work>
:/
<Sonarpulse-Work>
why not though?
<Sonarpulse-Work>
I just had to add more parens for chained or (at least I think that was the problem)
<Sonarpulse-Work>
and I am disgruntled about it
<Sonarpulse-Work>
of course, there can be a compatability grace period
ThatDocsLady has quit [(Ping timeout: 250 seconds)]
arjen-jonathan has quit [(Ping timeout: 265 seconds)]
<NixOS_GitHub>
[nixpkgs] FRidh pushed 1 new commit to master: https://git.io/v1ZtB
<NixOS_GitHub>
nixpkgs/master 9e45381 Frederik Rietdijk: pythonPackages.acoustics: init at 0.1.2
<Profpatsch>
Moving /nix/store/g9xd0kg1vq0bbrgfzh2k0krxgwpx526d-zlib-1.2.8/lib/libz.a to /nix/store/838xbbkm3qn31nc23qgn6k3p8vwpkkf8-zlib-1.2.8-static/lib/libz.a
<Profpatsch>
rmdir: failed to remove '/nix/store/g9xd0kg1vq0bbrgfzh2k0krxgwpx526d-zlib-1.2.8/lib': Directory not empty
takle has quit [(Remote host closed the connection)]
<Dezgeg>
aha, so they are in zlib.static then
<Profpatsch>
Dezgeg: Uh?
<Profpatsch>
This is strance.
<Profpatsch>
*ge
<Dezgeg>
nix-build -A zlib.static; find result-static/
<Profpatsch>
I thought nix-build creates links to all outputs by default.
<Profpatsch>
Did this change?
<Dezgeg>
I think it's always done that
goibhniu has quit [(Ping timeout: 248 seconds)]
ibrahims_ has joined #nixos
ibrahims has quit [(Ping timeout: 246 seconds)]
yodeler_ has joined #nixos
yodeler has quit [(Ping timeout: 250 seconds)]
alx741 has joined #nixos
pi3r has joined #nixos
ibrahims_ has quit [(Read error: Connection reset by peer)]
redmq has joined #nixos
jensens has quit [(Ping timeout: 260 seconds)]
katyucha has joined #nixos
katyucha_ has joined #nixos
rly has joined #nixos
<pierron>
Sonarpulse-Work: I won't be able to look at the pull request for the next 2 weeks.
eacameron has joined #nixos
eacameron has quit [(Remote host closed the connection)]
eacamero_ has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
<NixOS_GitHub>
[nixpkgs] peti pushed 4 new commits to master: https://git.io/v1Z2V
<NixOS_GitHub>
nixpkgs/master d80d927 Peter Simons: LTS Haskell 7.11
<NixOS_GitHub>
nixpkgs/master 10b832e Peter Simons: hackage-packages.nix: automatic Haskell package set update...
<NixOS_GitHub>
nixpkgs/master 880ae3f mingchuan: haskellPackages: fix brick and vty_5_13...
oida has joined #nixos
<NixOS_GitHub>
[nixpkgs] peti pushed 1 new commit to master: https://git.io/v1ZaN
<NixOS_GitHub>
nixpkgs/master 297ea75 Peter Simons: haskellPackages.ghc-core: Revert "0.5.6 -> 2012-12-15"...
eacameron has quit [(Read error: Connection reset by peer)]
eacamero_ has joined #nixos
eacameron has joined #nixos
eacameron has quit [(Remote host closed the connection)]
<Sonarpulse-Work>
pierron: thanks for the heads up
<Sonarpulse-Work>
....I'm really tempted to just merge it then
<Sonarpulse-Work>
:)
<Sonarpulse-Work>
(once travis does the OK)
eacameron has joined #nixos
eacamero_ has quit [(Read error: Connection reset by peer)]
eacamero_ has joined #nixos
<Sonarpulse-Work>
btw, we seem to have enough pull requests that travis's pull request tab is borked
<Sonarpulse-Work>
anybody know a better way to cancel builds?
<Sonarpulse-Work>
I suspect my old force-pushs are needlessly crowding the queue
<NixOS_GitHub>
[nixpkgs] garbas opened pull request #20844: common update script (do not merge, PR opened for comments) (master...updateScript) https://git.io/v1ZrY
herzmeister has quit [(Quit: Leaving)]
herzmeister has joined #nixos
simendsjo has joined #nixos
eacamero_ has quit [(Remote host closed the connection)]
<garbas>
gchristensen: ^^^ the PR you talked me into doing :)
eacameron has quit [(Read error: Connection reset by peer)]
glines has quit [(Ping timeout: 258 seconds)]
LysergicDreams has joined #nixos
simendsjo has quit [(Read error: Connection reset by peer)]
takle has joined #nixos
simendsjo has joined #nixos
<Profpatsch>
Dezgeg: I wonder where nix[pkgs] decides that the .a lands in the output static.
takle has quit [(Ping timeout: 245 seconds)]
NEo4v8_ has joined #nixos
mkoenig has quit [(Remote host closed the connection)]
pstn has quit [(Remote host closed the connection)]
pikajude has quit [(Ping timeout: 250 seconds)]
zagy has joined #nixos
amarsman has quit [(Ping timeout: 256 seconds)]
NEo4v8_ has quit [(Ping timeout: 268 seconds)]
pikajude has joined #nixos
mkoenig has joined #nixos
simendsjo has quit [(Remote host closed the connection)]
takle has joined #nixos
jmiven has quit [(Quit: co'o)]
jmiven has joined #nixos
<mkoenig>
Skype does not work anymore with XMonad since b09435e. Anyone same problem or is it me?
<Ralith>
in what way does it not work?
<mkoenig>
It opens either no window or an empty one.
zagy has quit [(Quit: Leaving.)]
rardiol has joined #nixos
<gchristensen>
garbas: yay!
<gchristensen>
garbas: <3 <3 <3
j`ey has joined #nixos
<Ralith>
mkoenig: weird; I'd try to bisect. It's been working fine on i3 for me, but I'm on 16.09
<Ralith>
and haven't updated in a while
<gchristensen>
sometimes I want to submit a mass rebuild just to get the # of workers to scale back up :P
<Ralith>
clang 3.9.1 should be coming along soon
glines has joined #nixos
<mkoenig>
yeah it IS weird. I have not even changed my xmonad.hs
<Ralith>
mkoenig: bisect nixpkgs, I mean
<Ralith>
I'd be surprised if the wm was responsible
<Ralith>
though... have you tried floating the window it opens?
<Ralith>
sometimes that unbreaks buggy apps
j`ey has left #nixos []
<NixOS_GitHub>
[nixpkgs] jgeerds pushed 2 new commits to master: https://git.io/v1ZMo
<NixOS_GitHub>
nixpkgs/master bfc2721 Jascha Geerds: pythonPackages.docker: Fix build for python3
<NixOS_GitHub>
nixpkgs/master d69a25d Jascha Geerds: pythonPackages.docker: Add myself as a maintainer
<magnetophon>
I'm trying to install NixOS on a 2010 imac, but X won't run. Been trying all day. I get "no screens found(EE)".
[0x4A6F]1 has quit [(Ping timeout: 258 seconds)]
mkoenig has quit [(Read error: Connection reset by peer)]
mkoenig has joined #nixos
<pmade>
I found a package that is missing a dependency and won't install. In my own package file, how do I add a dependency so that it will be available before this package tries to build?
<avn>
magnetophon: what lspci tell about video?
glines has quit [(Ping timeout: 260 seconds)]
glines has joined #nixos
<magnetophon>
avn: it's a HD 4670. does that answer your question
<pmade>
Something like the inverse of propagatedBuildInputs
<avn>
magnetophon: what remain in Xorg.log?
<magnetophon>
avn: I've not been able to get a log.
<Ralith>
pmade: you need to override that package to add the missing dep; you should consider patching the package directly and making a pull request to nixpkgs so everyone benefits
<avn>
magnetophon: why it not write log? Or you just haven't machine nearby now?
<pmade>
Ralith: I'm going to look at it. It's an automatically generated package (Haskell/hackage) and the upstream package is missing the dep.
zagy has joined #nixos
<savanni>
Hey, I have a really quick thing that I need to figure out.
<pmade>
The build actually succeeds but the test phase fails. I might just figure out how to turn that off.
<magnetophon>
avn: /var/log/ only has: faillog, journal/, lastlog, slim.log and wtmp.
<Ralith>
pmade: ah, just add doCheck = false; to the package, iirc
<savanni>
I have a derivation that I'm trying to make that is based off of an existing library, and that library is not present in nixpkgs. By itself, not a problem.
<Ralith>
pmade: maybe make a PR upstream too
<pmade>
Ralith: will do. Thanks.
<savanni>
However, the library won't compile due to a conflicting function name with glib.
<avn>
magnetophon: idk, bat be now logs go journalctl?
zagy has quit [(Client Quit)]
<savanni>
So, I've created a patch that renames the function in the library. How do I include the patch file in the derivation?
<Ralith>
not actually sure if there are any unusual issues with disabling tests for an auto-generated haskell package, but I doubt you're the first one to need it
<Ralith>
savanni: put the patch file next to the .nix file that defines the package, and then just refer it in the patches array
josdotso has joined #nixos
<avn>
magnetophon: does jaournalctl show something?
<gchristensen>
patches = [ ./my-patch.patch ];
<magnetophon>
avn: journalctrl gives the "no screens found" with the open drivers, and a segfault withthe closed drivers
<savanni>
Hmm... @gchristensen Shouldn't I see something in the output before the build script starts?
<Ralith>
savanni: also, you can probably delete the buildCommand stuff entirely, nixpkgs handles autotools automatically
<savanni>
Yeah, I added that in when I was trying to diagnose the problem.
<Ralith>
(and your --prefix is bad)
<savanni>
Also, patches start applying when I don't have the buildCommand stuff. :)
herzmeister has quit [(Quit: Leaving)]
<Ralith>
it may be clobbering the patch phase anyway
<Ralith>
just get rid of it
herzmeister has joined #nixos
zagy has joined #nixos
<Ralith>
normally if you're not using autotools you'd want to override build*Phase*, not buildCommand
<NixOS_GitHub>
[nixpkgs] FRidh pushed 10 new commits to master: https://git.io/v1ZH6
<NixOS_GitHub>
nixpkgs/master 49a313c Frederik Rietdijk: pythonpackages.lti: rename to PyLTI and remove overriding...
<NixOS_GitHub>
nixpkgs/master d3f3ad3 Frederik Rietdijk: pythonPackages.pyjwt: does not need pytest 2.7
<NixOS_GitHub>
nixpkgs/master a8379a1 Frederik Rietdijk: pythonPackages.pytestcache: disable tests
<savanni>
Yeah. Unfortunately, my patch doesn't apply yet. Still working on it.
zagy has quit [(Client Quit)]
AllanEspinosa has joined #nixos
newhoggy2 has quit [(Remote host closed the connection)]
newhoggy2 has joined #nixos
<savanni>
Got it. Had to modify the patch file to take into account where the archive would decompress to.
<magnetophon>
avn: I also get "check the log file at /dev/null for more info"
akaWolf has quit [(Ping timeout: 248 seconds)]
<magnetophon>
avn: also, I'm pretty sure lightdm is the default dm, and I have't configured another one, yet I get messages and log files about slim.. ??????
clever has joined #nixos
<avn>
magnetophon: nixos/modules/services/x11/xserver.nix add option for logfile with default /dev/null (as it hardcoded now)
<NixOS_GitHub>
[nixpkgs] vbgl pushed 1 new commit to master: https://git.io/v1Z73
<NixOS_GitHub>
nixpkgs/master 9c87884 Vincent Laporte: merlin: 2.5.2 -> 2.5.3
redmq has quit [(Ping timeout: 260 seconds)]
xadi has quit [(Ping timeout: 250 seconds)]
Itkovian has joined #nixos
newhoggy2 has quit [(Remote host closed the connection)]
newhoggy2 has joined #nixos
<magnetophon>
avn: wait, are you saying it's normal that the logs are sent to dev/null? And the way around it is to edit xserver.nix
<magnetophon>
?
newhoggy2 has quit [(Remote host closed the connection)]
newhoggy2 has joined #nixos
newhoggy2 has quit [(Remote host closed the connection)]
newhoggy2 has joined #nixos
goibhniu has joined #nixos
newhoggy2 has quit [(Remote host closed the connection)]
<Ralith>
magnetophon: no, normally X logs to the systemd journal
newhoggy2 has quit [(Remote host closed the connection)]
newhoggy2 has joined #nixos
<Ralith>
the logfile is set to /dev/null because it's redundant to the journal logging
zagy has joined #nixos
newhoggy2 has quit [(Remote host closed the connection)]
<avn>
Ralith: as we see, not all hit journal (or it not enough for troubleshouting), would be nice to have ability to direct everything to old fashion file
<NixOS_GitHub>
[nixpkgs] FRidh pushed 2 new commits to master: https://git.io/v1Z54
<avn>
but I think would be nice to have full (file) log
<Ralith>
avn: this is the full log
NEo4v8_ has quit [(Ping timeout: 250 seconds)]
<Ralith>
why are you still stuck on that
<Ralith>
magnetophon: your video driver appears to have crashed
<Ralith>
try a different one
mkoenig has quit [(Ping timeout: 240 seconds)]
<magnetophon>
avn: ah, sorry, I was booted into the wrong generation. I'll try another one with the open drivers. one sec
tvon has quit [(Quit: Peace out, y'all!)]
<Ralith>
magnetophon: in these situations it's often a good idea to start with something very conservative like vesa, get your system usable, and then experiment with faster drivers
<magnetophon>
Ralith I tried vesa, but that didn't get me to X either
<Ralith>
likely for a different reason, then
<Ralith>
have you pasted your configuration.nix anywhere yet?
<magnetophon>
Ralith it's almost empty
mkoenig has joined #nixos
<magnetophon>
and no, I haven't pasted it yet. will do
newhoggy2 has quit [(Remote host closed the connection)]
<magnetophon>
unrelated: the keyboard often doesn't work at boot time, so I can't choose a working generation. when I unplug the power for a while and boot again, it works. trying to be quick for you guys, I waited too short before replugging three times now, making me much sower in the end......
newhoggy2 has joined #nixos
<avn>
magnetophon: no rush. If anything go slow, I take look tomorrow morning
<gchristensen>
joachifm: thank you for tagging me
<mkoenig>
Ok, now I have an empty floating Skype window
NEo4v8_ has joined #nixos
newhoggy2 has quit [(Ping timeout: 260 seconds)]
<magnetophon>
avn: Ralith: http://pastebin.com/FwhNgK3n this is with no driver specified, and manually adding nomodeset to the boot line. (otherwise I don't get to the login screen)
<gchristensen>
offlinehacker: good issue
<pierron>
Sonarpulse-Work: would it be possible to make new PR for the debug instrumentation, only?
mkoenig has quit [(Ping timeout: 260 seconds)]
<avn>
magnetophon: `Dec 01 21:36:16 mommynix kernel: [drm:radeon_init [radeon]] *ERROR* No UMS support in radeon module!`
<avn>
so not sure, if radeon driver works w/o KMS
cpennington has quit [(Remote host closed the connection)]
qryantm has joined #nixos
mkoenig has joined #nixos
<magnetophon>
avn: I think I read somewhere that that's related to the nomodeset that I had to manually add to the boot params. (I've been reading xorg related forum posts all day, so I'm gtting a bit confused)
<Ralith>
magnetophon: is xf86-video-ati the current AMD driver? I vaguely recall it having been deprecated in favor of something newer
rly has quit [(Ping timeout: 265 seconds)]
<magnetophon>
avn: but if I don't do the nomodeset, I don't get to the boot screen. (at least not visually. I *can* log in blindly and reboot).
<Ralith>
avn's idea sounds plausible too
<Ralith>
magnetophon: that's conceivably due to issues in the radeon kernel module; what happens when you use vesa?
<magnetophon>
Ralith, iirc there is amdgpu which is newer, but that doesn't work either
<Ralith>
might not support your gpu if it's old
<magnetophon>
Ralith I don't get X with vesa either. one sec, I'll post a log
<Ralith>
really what you should do is find out what driver people who have your hw working are using
<magnetophon>
Ralith: yeah, I think I read that amdgpu doesn't support my card.
hiratara has quit [(Ping timeout: 258 seconds)]
<avn>
Ralith: is old radeon, so it `radeon` not `amdgpu`
josdotso has joined #nixos
<Sonarpulse-Work>
pierron: yes
redmq has quit [(Ping timeout: 248 seconds)]
<Sonarpulse-Work>
pierron: what I have right now is an alternate approach, not just one less commit
<Sonarpulse-Work>
but I was thinking of that anyways
<magnetophon>
Ralith: (re: find the right driver) true, but I haven't found a definitive answer. So no use in posting a log with vesa then?
<Sonarpulse-Work>
with the PR open, anybody that actually works on the stdenvs (i.e. not me) can just merge it if they need it
<Sonarpulse-Work>
rather than slamming their computer shut :)
<Ralith>
magnetophon: I can look at it if you like, but my best guess at this point is that there's a fatal bug in the radeon kernel module on your hw, so you might try disabling that somehow
<magnetophon>
Ralith wouldn't setting vesa as the driver accomplish that?
<gchristensen>
the classic problem with signed commits is you have to use GPG :)
<kmicu>
(Which is not a problem if you already use GnuPG to sign emails, encrypt files and log into remote servers ( ͡~ ͜ʖ ͡°))
AllanEspinosa has joined #nixos
<gchristensen>
kmicu: LOL
<gchristensen>
yes, however
glines has joined #nixos
mizu_no_oto has joined #nixos
goibhniu has joined #nixos
mkoenig_ has joined #nixos
mkoenig has quit [(Read error: Connection reset by peer)]
<kmicu>
(and actually, right now, to verify the new release of coreutils with ‘gpg --verify coreutils-8.26.tar.xz.sig’ ;)
<gchristensen>
for verifying signatures on packages, usually what people want is gpgv
<magnetophon>
avn, Ralith: OK, took a while, but I have a ssh-paste workflow setup. here's the log with ati: http://nixpaste.lbr.uno/kRCKTCau
goibhniu has quit [(Remote host closed the connection)]
<kmicu>
yey “ls now aligns quoted items with non quoted items, which is easier to read, and also better indicates that the quote is not part of the actual name.”
<Ralith>
magnetophon: doesn't look different from the last one; you can see the "drm report modesetting isn't supported" message, which isn't classed as an error but sure sounds fatal
<gchristensen>
ooh fancy, kmicu
goibhniu has joined #nixos
<gchristensen>
I wish we could use gpg in fetchurl and friends, but sadly we cannot
<gchristensen>
bootstrapping gets really complicated as far as I can tell
<Ralith>
magnetophon: ah, I've been a bit confused because all your log pastes contain historical data as well as current
<Ralith>
magnetophon: "V_BIOS address 0x0 out of range" seems like the problem with vesa; no idea what to make of that, googling suggests that people fix it by using a more suitable driver
<Ralith>
magnetophon: I'd focus on trying to apply that radeon module workaround
<magnetophon>
Ralith: OK, but it seems neither of us has an idea on where to start with that, right?
amyers has quit [(Ping timeout: 245 seconds)]
<viric>
mh suddenly icu4c does not build for nixos in 16.09
<clever>
magnetophon: any virtual machine stuff going on?
<Ralith>
magnetophon: if you can work out how to pass the appropriate option on module load, you can probably stuff the post-boot commands into a systemd unit that sequences itself before X
k2s has joined #nixos
<magnetophon>
clever: nope
k2s has quit [(Client Quit)]
<clever>
magnetophon: boot.kernelModules takes a list of modules to load on bootup
<clever>
this is meant to configure things linked into the kernel (so modprobe isnt possible), but it also works against loadable modules i believe
slack1256 has quit [(Remote host closed the connection)]
AllanEspinosa has joined #nixos
zagy has quit [(Client Quit)]
<Ralith>
well, work that out one way or another, then add a systemd unit with Before=display-manager.service that switches the options back
<offlinehacker[m]>
tnx
<mkoenig>
Ralith: it seems that not every commit is pushed to hydra...
<Ralith>
I wouldn't know anything about that
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
Itkovian has joined #nixos
<magnetophon>
Ralith, clever: lemme see if I understand correctly: step 1) according to that arch forum link Ralith gave me, I need to pass "options radeon modeset=0". I thik I read somewhere that is the same as passing nomodeset to the boot params. correct?
<Ralith>
I have no idea whether that's the same or not
<Ralith>
I'm inclined to suspect not
<Ralith>
you need to pass modeset=0 to the radeon module when it's first loaded, somehow
<magnetophon>
Ralith according to this link, I can do that in the boot params
<magnetophon>
hmmm, I'm running into the "non-working keyboard in uefi" again. need to unplug and wait for a bit....
civodul has quit [(Quit: ERC (IRC client for Emacs 25.1.1))]
<Ralith>
that looks credible, and also suggests that clever's idea is probably viable
<magnetophon>
Ralith which one?
qryantm has joined #nixos
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
<magnetophon>
Ralith: I copied over the "bind" thing he mentioned. is that what you mean?
<Ralith>
don't do that
<Ralith>
he was just giving a random example
<magnetophon>
oh
<magnetophon>
I already did
<Ralith>
god knows what it'll do to your system
<Ralith>
kernelParams = [ "radeon.modeset=0" ];
<magnetophon>
right
eacameron has joined #nixos
<magnetophon>
Ralith: OK, that get's me to a login prompt again. Now I should be able to just manually run the other commands from that arch page, right?
aj__ has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<Ralith>
magnetophon: yes, that'd be a good next step
<Ralith>
run those commands and restart display-manager.service
<Ralith>
if that works, then we can ensconce them in your configuration
<magnetophon>
Ralith: modprobe -r -v radeon drm gives me a fatal error: module drm is i use...
<magnetophon>
btw: I changed the videodriver to radeon. is that right?
<Ralith>
magnetophon: lsmod |grep ^drm
<Ralith>
and yes
<magnetophon>
Ralith: that gives 2 lines: drm_kms_helper, and drm
<magnetophon>
do you need the full output?
<Ralith>
magnetophon: the rightmost column on the "drm" line is what you need to add to your modprobe -r command
<Ralith>
without the commas
<Ralith>
with spaces instead of commas, rather
<magnetophon>
ok, no output on that comand, which I assume is good. now I can also remove drm. I'll move on with the rest of the arch commands
<magnetophon>
Ralith: aaaand the display hangs again, when I insert the radeon module. ssh still works.
<magnetophon>
Ralith: suggestions?
mkoenig has quit [(Read error: Connection reset by peer)]
glines has quit [(Ping timeout: 250 seconds)]
* Ralith
shrug
<Ralith>
complain to AMD? :P
<magnetophon>
tbc: when I run "modprobe -v radeon modeset=1", I get 7 "insmod" lines, the last one being: "insmod /path/to/radeon.ko.xz modeset=0 modeset=1" and then a line with just "_".
<magnetophon>
Ralith: I'll write a letter to Apple. I'm sure they'll get right on it.
<Ralith>
you could experiment with older versions of the driver and/or kernel at large
mkoenig has joined #nixos
<magnetophon>
Ralith: anyway, again many thanks for your patience. I think I'll try a few other distros with other drivers/kernels. Eventually I'd like to get NixOS running though, as it has served me well for about 2 years now.
<Ralith>
magnetophon: fwiw, nix makes it very easy to test different kernels with minor config changes
<Ralith>
looks like it goes back to 3.10
<magnetophon>
yeah, that might actually be easier.
<Ralith>
not sure how that interacts with driver versions though
thc202 has quit [(Ping timeout: 260 seconds)]
<eacameron>
what's the difference between || and or?
<jeaye>
eacameron: In C++, nothing.
<jeaye>
Woops, wrong channel.
<eacameron>
jeaye: Lol. What about in nix?
<jeaye>
Don't know. Thought I was in ##c++.
<magnetophon>
Ralith: and I'd try without the kerneparams again, right?
<Ralith>
magnetophon: that's the ideal, I suppose, but that thread was from 2013; it might be the case that you need an old kernel *and* the workaround
amyers has joined #nixos
<magnetophon>
Ralith, ok, kernel 3.18 with the workaround gives a hanging display again (I forgot to remove the kernelParams line, so I figured I'd try the workaround first.) Will try without params now.
chattered has quit [(Quit: ERC (IRC client for Emacs 25.1.1))]
kampfschlaefer has quit [(Ping timeout: 258 seconds)]