<hodapp>
how might I run a command as part of setting a container up? I am not really sure from reading whether or not, say, postStartScript is something I should be editing
pradeepch has joined #nixos
derjohn_mob has quit [(Ping timeout: 240 seconds)]
kier has quit [(Remote host closed the connection)]
sellout- has joined #nixos
kaychaks has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
derjohn_mob has joined #nixos
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
eacameron has joined #nixos
mizu_no_oto has joined #nixos
pmade has left #nixos ["Using Circe, the loveliest of all IRC clients"]
ppppp5 has quit [(Remote host closed the connection)]
ppppp5 has joined #nixos
<Rotaerk>
hmm, nixpkgs/pkgs/development/haskell-modules/hackage-packages.nix is supposedly auto-generated, but what's the source?
<Rotaerk>
I don't see how some of what's in it could be generated based on the cabal files alone
<Rotaerk>
such as BerkeleyDBXML's {inherit (pkgs) db; dbxml = null; inherit (pkgs) xercesc; inherit (pkgs) xqilla;}
<Rotaerk>
particularly, how it would know to pass in null for dbxml when it's passing in a package for the others; nothing in the cabal file seems to distinguish them
erictapen has joined #nixos
fiddlerwoaroof has quit [(Ping timeout: 260 seconds)]
<Rotaerk>
is there a script somewhere that generates the hackage-packages.nix? or is it just hackage2nix with all default args?
fiddlerwoaroof has joined #nixos
glenn has joined #nixos
glenn has quit [(Remote host closed the connection)]
glenn has joined #nixos
mrkgnao has joined #nixos
e3d3 has joined #nixos
mkoenig has quit [(Remote host closed the connection)]
indi_ has joined #nixos
e3d3 has left #nixos ["rcirc on GNU Emacs 25.1.1"]
Kingsquee has quit [(Ping timeout: 255 seconds)]
Kingsquee has joined #nixos
ryantrinkle has joined #nixos
glenn has quit [(Remote host closed the connection)]
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
eacameron has quit [(Remote host closed the connection)]
eacameron has joined #nixos
<mrkgnao>
so Intero isn't working for me because it can't find a .so that comes from a package that is in my nix-shell environment
<mrkgnao>
(I'm running emacs from the nix-shell)
eacameron has quit [(Ping timeout: 240 seconds)]
eacameron has joined #nixos
<NixOS_GitHub>
[nixpkgs] benley opened pull request #22080: exercism: Broaden platforms to include darwin, etc (master...exercism) https://git.io/vMFKm
mguentner has quit [(Quit: WeeChat 1.6)]
eacameron has quit [(Ping timeout: 240 seconds)]
k0001 has quit [(Quit: Connection closed for inactivity)]
hexagoxel has quit [(Ping timeout: 240 seconds)]
eacameron has joined #nixos
mguentner has joined #nixos
hexagoxel has joined #nixos
eacameron has quit [(Ping timeout: 276 seconds)]
sellout- has quit [(Quit: Leaving.)]
glenn has joined #nixos
mizu_no_oto has joined #nixos
stepcut has quit [(Remote host closed the connection)]
eacameron has joined #nixos
<NixOS_GitHub>
[nixpkgs] benley pushed 2 new commits to master: https://git.io/vMFis
<NixOS_GitHub>
nixpkgs/master be1b5c2 Benjamin Staffin: exercism: Broaden platforms to include darwin, etc
<NixOS_GitHub>
nixpkgs/master b2b32ea Benjamin Staffin: Merge pull request #22080 from benley/exercism...
stepcut has joined #nixos
glenn has quit [(Remote host closed the connection)]
glenn has joined #nixos
georges-duperon has quit [(Ping timeout: 252 seconds)]
glenn has quit [(Remote host closed the connection)]
stepcut has quit [(Remote host closed the connection)]
glenn has joined #nixos
glenn has quit [(Read error: Connection reset by peer)]
mguentner2 has joined #nixos
mguentner has quit [(Ping timeout: 258 seconds)]
Wizek_ has joined #nixos
stepcut has joined #nixos
<swflint>
Hey, I've got a fresh install, and I'm getting the following error: http://i.imgur.com/pOFjiz3.jpg What can I do?
glines has quit [(Ping timeout: 240 seconds)]
nathan7 has quit [(Ping timeout: 255 seconds)]
nathan7 has joined #nixos
filterfish_ has quit [(Ping timeout: 245 seconds)]
AllanEspinosa has quit [(Ping timeout: 240 seconds)]
<Benrob0329[m]>
Did you create a new user?
glenn has joined #nixos
<swflint>
I did define a user, yes.
<Benrob0329[m]>
I would disable Lightdm and see if SliM starts
nathan7 has quit [(Ping timeout: 240 seconds)]
nathan7 has joined #nixos
<swflint>
Okay. I can try that.
glines has joined #nixos
koenvg_ has joined #nixos
glenn has quit [(Remote host closed the connection)]
glenn has joined #nixos
indi_ has quit [(Remote host closed the connection)]
indi_ has joined #nixos
filterfish has joined #nixos
indi_ has quit [(Ping timeout: 240 seconds)]
<koenvg_>
Hey! Just getting into NixOS, and I'm really liking it. One small issue I have: my ~/.config/gtk-3.0/gtk.css don't seem to be taken into account. Any idea why?
<benley>
oh boy the mysteries of gtk theming
<benley>
I have never been able to make heads or tails of that
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
<NixOS_GitHub>
[nixpkgs] svanderburg pushed 1 new commit to master: https://git.io/vMbv7
<NixOS_GitHub>
nixpkgs/master 4f7fe27 Sander van der Burg: titaniumenv: use Node.js 4.x for the CLI tools
<mrkgnao>
does anyone have Intero working on nixos?
<symphorien>
how can I override the priority of a package in the system environment ? I try to have netcat openbsd instead of netcat-gnu so I put netcat-openbsd.overrideDerivation(old: { meta.priority = "5"; }) in systemPackages but I get the message "cannot coerce a function to a string".
<mrkgnao>
old.meta.priority maybe?
<mrkgnao>
just guessing, I don't exactly know what's going on
<symphorien>
same result :(
derjohn_mob has joined #nixos
<symphorien>
I also tried with 5 as an int instead of a string, and with meta = { priority = 5 ; };
<NixOS_GitHub>
[nix] edolstra pushed 2 new commits to master: https://git.io/vMbfW
<NixOS_GitHub>
nix/master 1102c77 Eelco Dolstra: shell.nix: Add a flag for using clang
<NixOS_GitHub>
nix/master e5641df Eelco Dolstra: Work around a bug in clang and older versions of gcc...
jsgrant- has quit [(Excess Flood)]
jsgrant- has joined #nixos
mudri has quit [(Ping timeout: 255 seconds)]
ilja_kuklic has quit [(Ping timeout: 240 seconds)]
fritschy has quit [(Remote host closed the connection)]
fritschy has joined #nixos
<NixOS_GitHub>
[nixpkgs] timbertson opened pull request #22086: buildPythonPackage: send pip output to stderr (master...pip-stderr) https://git.io/vMbJ9
<NixOS_GitHub>
[nixpkgs] timbertson closed pull request #22015: buildPythonPackage: accept developmentPrefix and send `pip` output to stderr (master...python-stderr) https://git.io/vMQvy
Biappi has quit [(Ping timeout: 258 seconds)]
<NixOS_GitHub>
[nixpkgs] edolstra pushed 1 new commit to release-16.09: https://git.io/vMbU3
<NixOS_GitHub>
nixpkgs/release-16.09 b26fa70 Eelco Dolstra: Revert "Use looser 9pfs caching in VM tests/builds"...
<ronny>
chpatrick: i beleive all that would be needed is such a parameter
avn has quit [(Ping timeout: 240 seconds)]
ixxie has joined #nixos
<pierron_>
garbas: It's your turn to visit London :P
indi_ has quit [(Remote host closed the connection)]
<pierron_>
garbas: I hope you will have more persons attending that what I had last time.
deepfire has quit [(Ping timeout: 264 seconds)]
<ixxie>
oh.... they use Nix at Mozilla :>
<ixxie>
nice
Wizek has quit [(Ping timeout: 240 seconds)]
Wizek_ has quit [(Ping timeout: 245 seconds)]
<NixOS_GitHub>
[nixpkgs] svanderburg pushed 1 new commit to master: https://git.io/vMbZY
<NixOS_GitHub>
nixpkgs/master de9a4f5 Sander van der Burg: titaniumenv: add 6.0 SDK + bump kitchensink testcase
avn has joined #nixos
pradeepch has joined #nixos
jluttine has joined #nixos
indi_ has joined #nixos
<gchristensen>
garbas: someone tells me they _just now_ realize 16.09 was out. perhaps we should put some sort of notice in old versions of nixpkgs or an update nagger or something
pradeepch has quit [(Ping timeout: 258 seconds)]
calvinx has joined #nixos
<LnL>
gchristensen: :o
<pierron_>
gchristensen: That's supposed we are capable of releasing in time.
<pierron_>
gchristensen: unless you want something which pulls a simple text file from nixos.org which contains the latest version number.
<pierron_>
Which I admit would be interesting for knowing the number of active installations.
deepfire has joined #nixos
<gchristensen>
or (if we assume they do nix-channel --update) by putting a change in there somehow
<gchristensen>
a change in nixpugs*
<gchristensen>
nixpkgs* (I can type, I swear)
<pierron_>
hum, sounds easily doable then.
zml has joined #nixos
<MoreTea>
So when will 16.09 not be maintained anymore?
<MoreTea>
Directly after the 17.03 release?
<MoreTea>
Or is there a period in which both 16.09 and 17.03 are supported?
cpennington has quit [(Remote host closed the connection)]
<gchristensen>
that is typically when people stop backporting patches to it, MoreTea, but if someone in the community wanted to backport stuff to 16.03 it'd be fine
deepfire has quit [(Ping timeout: 255 seconds)]
<MoreTea>
Ideally, we should communicate clearly that some release is not supported anymore after $DATE, and provide a (small) window in which we'd support both $LATEST_STABLE and $LATEST_STABLE-1. A window of ~ 1 month should be good enough
xadi has quit [(Quit: Leaving.)]
<zimbatm[m]>
ideally we have enough man power to support two LTS releases like Ubuntu
<NixOS_GitHub>
[nixpkgs] svanderburg pushed 1 new commit to release-16.09: https://git.io/vMbcQ
<NixOS_GitHub>
nixpkgs/release-16.09 2f77bef Sander van der Burg: titaniumenv: add 6.0 SDK + bump kitchensink testcase...
<gchristensen>
zimbatm[m]: in reality though?
<gchristensen>
MoreTea: I _think_ I could commit to applying patches to 16.09 for another month
<zimbatm[m]>
in reality we don't have enough manpower to even support one release properly (yet)
<MoreTea>
I guess that will only become a reality when businesses want to pay $$$ for long term support
<joepie91>
I would suggest limiting old-release support, for now, to "backporting of security patches only", so that it can be rolled in with the existing vulnerability roundups
<joepie91>
and only for a limited duration, 1 month is probably enough
<joepie91>
given the technical characteristics of NixOS
<zimbatm[m]>
yeah that would be nice
<joepie91>
(namely: it's not likely to blow up in your face)
<MoreTea>
Yep, this one month overlap will allow people to switch to the newer stable.
<gchristensen>
I'll make sure to ping you two on those roundups ;)
<MoreTea>
hehe, I'll try to help when possible ;)
<gchristensen>
joepie91 sent me a bunch of ideas on how to nag users, I'm thinking the most likely option would be adding a like ... `builtin.trace` somewhere important in the old nixpkgs trees so they might have a chance of seeing it during a rebuild
deepfire has joined #nixos
<gchristensen>
"most likely option" as in most likely to be done before 17.03
<goibhniu>
gchristensen: FWIW, I've been working on a little dashboard icon to show when a new channel is available ... can also easily show stable releases
<goibhniu>
dashboard/system tray
<gchristensen>
that could be good
<joepie91>
goibhniu: yeah, that sounds like exactly what we need [in graphical environments]
<gchristensen>
we'll likely want to patch something like this in to even really really old releases
<gchristensen>
goibhniu: I have a way for you to know when updates contain security patches ...
<MoreTea>
gchristensen, that would be the nice way. The other one would be to do something like `allowUnfree`; abort unless `yesIWantToUseUnsupportedRelease = "16.09";`
<goibhniu>
gchristensen: sweet!
<gchristensen>
pierron_: ^ opinions?
<MoreTea>
It's probably quite easy to skip over a `builtins.trace`
<joepie91>
MoreTea: keep in mind that NixOS isn't just used in headless environments :P
<joepie91>
and documentation doesn't solve the problem of "people have X things to do, and a daily check of whether there's a new NixOS version isn't one of them"
<pierron_>
gchristensen: One of the use case if for reproducibility of research programs, and I would defintely see NixOS fill this gap better than VM, to the extend that CPU instructions are stable enough.
<pierron_>
s/NixOS/Nix/
<gchristensen>
pierron_: oh yes agreed
<pierron_>
gchristensen: I do not think we should prevent users from using deprecated versions at any time.
<gchristensen>
hmm yeah
<gchristensen>
and having to change the code to permit it is annoying
lonokhov has quit [(Quit: leaving)]
<gchristensen>
but a trace would be appropriate I believe
romildo has joined #nixos
<pierron_>
I think we need a way to request user input in a yes/no format from Nix evaluation.
<pierron_>
or maybe yes/throw format.
<MoreTea>
Hm. My main problem is that these traces must be observed during the nix-building, but if you do a nix-build, and go grab some coffee, you'll never see it.
<joepie91>
gchristensen: having to change the code to permit it would be annoying, yes, but that's the point :)
<goibhniu>
thanks gchristensen!
<MoreTea>
This could be solved by having a `builtin.addSummaryWarning` that would print warnings just before nix-build exists
<pierron_>
joepie91: but then it is no longer reproducible.
<gchristensen>
breakfast time, back soon. think on this :)
<joepie91>
gchristensen: your suggestion being to have this warning the moment a new release comes out ("This release is outdated and will stop being supported on $date, please update to $newVersion ASAP"), and when the old version goes EOL after a month or so, to change it to an EOL notice ("This release is outdated and no longer receives security updates; you should update to $newVersion immediately")?
<pierron_>
At least we need to improve the user readable outputs from Nix.
<LnL>
a single user install on otherr distros uses nixpkgs-unstable my default
calvinx has quit [(Ping timeout: 240 seconds)]
<MoreTea>
pierron_, indeed.
calvinx has joined #nixos
<MoreTea>
So this might be too late to fix for 17.03, given that that would probably require some changes to nix
<gchristensen>
ok so the other thing is we can start with something not very good and make it better, we have the technology :) but we need to start somewhere
<NixOS_GitHub>
[nix] edolstra pushed 2 new commits to master: https://git.io/vMblc
<pierron_>
before --> right before the programs start with the init system.
<MoreTea>
joepie91, that issue is orthogonal to this one
<MoreTea>
but a valid one nonetheless
<joepie91>
MoreTea: right, but that's why I'm asking "computing whether a nixpkgs is out of date" seems out of scope for the issue of old NixOS versions
<joepie91>
:P
<joepie91>
er, asking,*
<joepie91>
pierron_: the goal isn't to notify the user of security issues; rather to notify them that a new release is available and/or that their current one is EOL
dramforever has joined #nixos
avn has quit [(Ping timeout: 248 seconds)]
<joepie91>
so that it only shows after services have been started is not a big deal in that context
<joepie91>
I do think that "refuse building an EOL NixOS without opt-in" is a next step to be considered for the moment when the transition period has ended, but that's a next step
<joepie91>
and probably more controversial
<joepie91>
so let's just start with warnings :P
<MoreTea>
I meant, having the infrastructure in place so that we could have a state machine for release lifecycles; unstable -> stable -> old_stable -> unsupported
<pierron_>
joepie91: I disagree, with the opt-in.
<pierron_>
joepie91: the user already opted-in by choosing the channel.
<joepie91>
MoreTea: I'm just not really seeing how this relates to "computing whether a nixpkgs is out of date" - a "timebomb" is precisely what you'd want to avoid due to reproducibility, so manually pushing the warning to older NixOS branches seems like a better solution to me than anything automated
<joepie91>
pierron_: no, they didn't
<joepie91>
pierron_: and like I said, this is likely to be more controversial, so this needs more discussion
<MoreTea>
it's not "computing" anything; it's just something that you'd commit to nixpkgs, and then gets on your machine when you do a `nix-channel --update`
<joepie91>
(at the time the user opted into using the channel, it was likely not EOL; when it goes EOL, that opt-in needs reconsideration)
<MoreTea>
I'm proposing to do exactly what you're talking about :)
<LnL>
yeah just adding a new commit to the release branch that's out of date sounds reasonable to me
<pierron_>
nix-channel --update is not trivially reversible from the command line, we should not break anything!
<joepie91>
MoreTea: right, your use of the term "compute" threw me off :P
JagaJaga has quit [(Ping timeout: 258 seconds)]
<LnL>
pierron_: nix-channel --rollback
<LnL>
:)
<joepie91>
MoreTea: but yeah, then I think we agree
<pierron_>
LnL: ok, then we should advertise this one.
<rly>
I would like an interface nix-undo myself.
<mrkgnao>
how do I fix "no space left on device" errors? I'm getting those with stack
<mrkgnao>
seems like the /run/user/uid tmpfs is getting full
<joepie91>
there are a bunch of things with further discussion in the comments, but overall I feel like this would be a *massive* step forwards compared to the current command set
phreedom has quit [(Ping timeout: 276 seconds)]
<joepie91>
in terms of intuitiveness and general making-sense-ness
<MoreTea>
It would be really interesting if we'd had some command analogue to `{bundler,npm,yarn} add <pkg>` that would parse a default.nix expression and do some magic
<joepie91>
there's a much more direct mapping between "things users want to do" and "commands there are"
<joepie91>
whereas right now you're usually left hunting obscure flags for 30 minutes
<MoreTea>
indeed
<MoreTea>
or even command name
<joepie91>
yep
<ixxie>
joepie91: that is pretty cool..... any ideas of when / whether this will be implemented?
<MoreTea>
AFAIK, it's already slowly being implemented
<joepie91>
ixxie: I'm not really sure. the gist has been sitting there for a while now (earliest revision is a year ago), and I've heard that some bits are starting to be adapted, but I don't know if there's a concrete milestone / issue thread for it, or whether discussion around all of it has completed yet
<joepie91>
I found the gist more or less by accident, I think it was referenced from an issue thread somewhere :P
<joepie91>
so I don't know much about the process behind it
indi_ has quit [(Remote host closed the connection)]
<ixxie>
maybe copumpkin does since they were an early commented on the issue?
xadi has joined #nixos
<ixxie>
btw has anybody ever tried to explore the possibility of configuring /sys/devices via NixOS modules?
mizu_no_oto has joined #nixos
Seichi has joined #nixos
avn has joined #nixos
georges-duperon has joined #nixos
<joepie91>
I should probably get to work...
deepfire has quit [(Ping timeout: 260 seconds)]
chpatrick has left #nixos []
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
xadi has quit [(Quit: Leaving.)]
jgertm has quit [(Ping timeout: 255 seconds)]
asymmetric has joined #nixos
<gchristensen>
pierron_: you'd see it if you ran a switch, but I see what you're saying that it isn't enough.
<gchristensen>
like if you do a nixos-rebuild boot
<pierron_>
gchristensen: yes, on the boot it will only appear a few fraction of seconds before changing screens,
<gchristensen>
what do you think the simplest possible method would be?
<NixOS_GitHub>
nix/master fa738e5 Eelco Dolstra: Revert "builtins.readFile: Put the references of the file, not those needed to realize the file, into the context"...
<NixOS_GitHub>
[nix] edolstra pushed 1 new commit to master: https://git.io/vMbVu
<gchristensen>
niksnut: I'd like you to take a look at https://github.com/NixOS/nixpkgs/pull/22096 (the rule I was given when I got merge rights was not do anything controversial ;) )
<MoreTea>
Sure. I'm hoping to provoke some responses from people.
<MoreTea>
I agree that the abort should probably disappear, but the problem then is that this becomes all very hard to discover.
<avn>
Folks, anyone have zfs pool on top of lvm/md1?
<jarlg>
jarlg: It seems like it'd be a sane default.. wonder what's the argument against? (People using other shells maybe?)
jsgrant has joined #nixos
<jarlg>
LnL: ^ ...
fritschy has quit [(Quit: Leaving.)]
<LnL>
gchristensen: does it cause a rebuild for that many things?
<gchristensen>
it is pretty deep ..
iyzsong has quit [(Quit: bye)]
<LnL>
right, it might rebuild git
<ToxicFrog>
avn: in general that's disrecommended compared to just letting zfs manage the devs itself; why do you ask?
<avn>
ToxicFrog: in middle of migration now ;)
<jophish>
regnat: sounds exciting, I'm really looking forward to it
<avn>
ToxicFrog: just found that random processed blocked on write to zfs
<clever>
avn: my laptop has zfs on lvm on luks
corngood has joined #nixos
<philipp[m]>
avn: I mean, it's nothing you'd want to do but I don't see that it's really harmfull. Do you have any problems?
<avn>
clever: any specific tunings?
<avn>
then processes unblock after 1-3 mins of waiting
<gchristensen>
<snark> one major tuning I've heard is "don't use zfs unless you have a ton of drives"
<Benrob0329[m]>
FireFox 51 has been released
<avn>
Actually I plan to kick lvm/md from under zpool, as soon I move root onto zfs, and see if it succesful booting
t42 has joined #nixos
<clever>
avn: havent messed with any of the tuning yet
<LnL>
avn: did your pool get filled up at some point?
<avn>
I think it block size on md1 (possible)
<clever>
avn: main reason i went with zfs on lvm, is because i wanted swap on the same luks, but i didnt want to deal with swap on a zvol
systemfault has joined #nixos
<avn>
clever: swap on zvol looks working
<clever>
avn: so the only thing lvm is doing is partitioning up a luks volume
<clever>
only 1 physical drive
<philipp[m]>
gchristensen: Seconded. Also you want to only use it with ecc ram, see "Scrub Of Death".
<LnL>
that's fine, but using and md device with zfs is probably not a good idea
bennofs1 has joined #nixos
<Benrob0329[m]>
Er, FF will be released today, sorry
<avn>
LnL: I think it should b filled on 90-95% after rsync of my film archive will be done, then I kill second last volume (vg0/video) and give this space to `tank volume`
<avn>
LnL: Just haven't spare disk to make simpler migration
<LnL>
filling up a pool that much can also cause performance issues
<avn>
LnL: it ~50% atm
<t42>
Hi, quick question: I'm currently logged in on root with slim+X+i3, and whenever I switch to a tty via ctrl+alt+F?, the system freezes and I can't get back to my X session. This worked before enabling the xserver. Does this ring an obvious bell for anyone?
<yorick>
is it me or does NIX_BUILD_HOOK have no effect?
<avn>
I move stuff from my ~250G volume onto zfs, then give space to zfs
<yorick>
NIX_BUILD_HOOK=nix/build-remote.pl nix-build '<nixpkgs>' -A i3status just buillds locally
asymmetric has quit [(Ping timeout: 264 seconds)]
<avn>
then I move my root, and cross fingers if it boot from zfs on next reboot ;)
<Benrob0329[m]>
> Hi, quick question: I'm currently logged in on root with slim+X+i3, and whenever I switch to a tty via ctrl+alt+F?, the system freezes and I can't get back to my X session. This worked before enabling the xserver. Does this ring an obvious bell for anyone?
<Benrob0329[m]>
t42 (IRC) Yes and no
<LnL>
avn: you can't defragment a zpool so you have to be a bit careful with filling it up above 80%
xadi has quit [(Quit: Leaving.)]
<Benrob0329[m]>
What graphics driver are you using?
<avn>
LnL: will it defragment on resilvering (later)?
derjohn_mob has quit [(Ping timeout: 258 seconds)]
<Benrob0329[m]>
And why did you log into X11 as root?
<LnL>
avn: don't know exactly, but I think it might not
<MoreTea>
avn, the only way to get it OK again is to destroy your pool, and import an backup
<MoreTea>
But it'll just reduce performance a bit.
<LnL>
the md device is probably causing more issues then this
<t42>
Benrob0329[m]: Because I'm early into the configuration and didn't get around to setting up my own user yet. Yes, I should do that. I switched to linuxPackages_4_8 because with 4.4 there was something about Skylake integrated graphics support lacking. This is an Optimus laptop, but I have not installed any nvidia drivers yet
<MoreTea>
I solved this by creating an additional filesystem, and set a reservation, so that the disk will never get too full
<LnL>
I just set a quota on the root pool
<Benrob0329[m]>
t42: go ahead and do both of those
<Benrob0329[m]>
t42: it should fix your problem
<Benrob0329[m]>
BTW you can do both on Install
avn has quit [(Ping timeout: 264 seconds)]
dbmikus has quit [(Quit: WeeChat 1.7)]
dbmikus has joined #nixos
<Benrob0329[m]>
t42: be sure to set ``` nixpkgs.config.allowUnfree = true;```
Biappi has joined #nixos
ryantrinkle has left #nixos []
<t42>
Benrob0329[m]: I enabled bumblebee and added "nvidia" to videoDrivers (and added a user). This might take a couple of reboots, thanks for the help! :)
pradeepch has joined #nixos
tvon has joined #nixos
<Benrob0329[m]>
t42: No problem
avn has joined #nixos
<avn>
LnL: my plans are degrade md1, then give freed disk to zpool as mirror and resilver
<avn>
then finally kill md1, and resilver again
simendsjo has joined #nixos
<avn>
`iostat 1` show me only read activity when stuff (including rsync) blocked
<philipp[m]>
t42: Do you use mesa or some proprietary drivers? I had similar issues with fglrx years ago.
asseadinho has quit [(Quit: Leaving.)]
simendsj` has joined #nixos
simendsj` has quit [(Client Quit)]
simendsjo has quit [(Remote host closed the connection)]
Seichi has quit [(Remote host closed the connection)]
Seichi has joined #nixos
pradeepch has quit [(Read error: Connection reset by peer)]
sellout- has joined #nixos
derjohn_mob has joined #nixos
simendsjo has joined #nixos
<swflint>
I've got the following error on a recent install when trying to run X. What can I do? http://i.imgur.com/zlb6uaa.jpg
avn has quit [(Ping timeout: 240 seconds)]
JagaJaga has joined #nixos
xadi has joined #nixos
sellout- has quit [(Remote host closed the connection)]
sellout- has joined #nixos
bennofs1 has quit [(Ping timeout: 255 seconds)]
Intensity has quit [(Changing host)]
Intensity has joined #nixos
deepfire has joined #nixos
osa1 has joined #nixos
pi3r has quit [(Ping timeout: 248 seconds)]
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<goibhniu>
hi swflint, what gpu have you got and how is it configured in NixOS?
ckauhaus has joined #nixos
<swflint>
Integrated, intel I believe. No direct config. This is on a laptop.
roconnor has joined #nixos
<goibhniu>
swflint: ah, you probably need services.xserver.videoDrivers = [ "intel"];
<goibhniu>
swflint: how are you trying to run x?
<swflint>
Okay.
<goibhniu>
you need a display manager on NixOS
<swflint>
Through the display-manager service.
<goibhniu>
cool
Avaq has quit [(Ping timeout: 255 seconds)]
avn has joined #nixos
<swflint>
Also, is it normal for a nix install to be really slow?
<Benrob0329[m]>
I would leave SliM enabled until your sure it works, then mess with it.
<goibhniu>
swflint: not at all
<goibhniu>
swflint: is it compiling stuff?
<swflint>
Not that I'm aware of.
<goibhniu>
how much RAM have you got?
jgertm has quit [(Quit: WeeChat 1.6)]
<goibhniu>
over 2G?
jgertm has joined #nixos
<swflint>
Just 2.
<Benrob0329[m]>
Swap?
<swflint>
Have 3g set aside as swap
<gchristensen>
swflint: what is your computer doing right now? does it have high load? (what does `uptime` say?) if you look at `top` are there processes taking up lots of CPU?
avn has quit [(Ping timeout: 276 seconds)]
<gchristensen>
swflint: are you running low on memory? (what does `free -m` say?)
<sphalerite>
swflint: I misunderstood that as using a 3G connection as a swap device. That would be... interesting
<NixOS_GitHub>
[nixpkgs] dtzWill opened pull request #22098: swift: init at DEVELOPMENT-SNAPSHOT-2017-01-23-a (master...feature/swift) https://git.io/vMbAa
eacamero_ has quit [(Remote host closed the connection)]
<dtz>
although linux-only for the moment, not sure if you're interested in having it work on darwin as well?
<LnL>
yes, I want it for a darwin package
avn has joined #nixos
<dtzWill>
oh, sorry :3. I've gotten pretty familiar with how the build and such works so can probably help make that happen.... but no darwin host to test on here O:).
rly has quit [(Ping timeout: 256 seconds)]
ckauhaus has quit [(Ping timeout: 240 seconds)]
<ToxicFrog>
gchristensen: ! I'm stealing that snippet for later
<gchristensen>
:)
<LnL>
dtzWill: no problem, I can test it
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
stepcut has quit [(Remote host closed the connection)]
Willi_Butz has quit [(Ping timeout: 248 seconds)]
stepcut has joined #nixos
morsicus1 has joined #nixos
stepcut has quit [(Remote host closed the connection)]
<swflint>
gchristensen: no iostat output.
Willi_Butz has joined #nixos
reinzelmann has joined #nixos
pi3r has joined #nixos
<corngood>
Are you running in a VM?
<swflint>
Nope. Laptop
<corngood>
Yeah just checking that it's bare metal.
<corngood>
This is where you start to see the upside of nixos :)
<swflint>
Thanks!
<swflint>
And yes, it is.
<philipp[m]>
Systemd 228 has a local root exploit that was fixed but not documented. Is nixos using that in any version? https://lwn.net/Articles/712367/
civodul has quit [(Remote host closed the connection)]
<gchristensen>
philipp[m]: 15.09 and 16.09 do not, not sure about 16.03
<clever>
eek, i really need to update my router!
<clever>
its got 228...
<gchristensen>
philipp[m]: 16.03 is also safe as well
<philipp[m]>
gchristensen that could have been nasty.
<ixxie>
has anybody ever considered extendix Nix to the init system? Something for NixOS like DMD is for Guix?
<gchristensen>
systemd does nicely
Seichi has quit [(Quit: Konversation terminated!)]
<simpson>
ixxie: Sure. It'll be a lot of work.
<ixxie>
simpson: I am sure it is, I am just wondering if anybody has toyed with the idea....
<simpson>
ixxie: Yes. It looked like a lot of work and I don't have time to commit to more a-lot-of-work projects.
nschoe has quit [(Quit: Program. Terminated.)]
<clever>
philipp[m]: potentialy related, at a glance, i think thats for imperatively made timers, which nixos has probably disabled?
<clever>
declarative or bust!
<ixxie>
simpson: makes sense
HappyEnte has quit [(Quit: ZNC 1.6.3 - http://znc.in)]
* clever
heads to bed
eacamero_ has joined #nixos
morsicus1 has quit [(Quit: Leaving.)]
<NixOS_GitHub>
[nixpkgs] copumpkin opened pull request #22103: apache-kafka service: change default brokerId to -1 (master...automatic-kafka-broker-id) https://git.io/vMNL2
<ixxie>
In the development of a highly modular codebase on NixOS, when does it make sense to use nix packaging as opposed to - say - git modules/subtrees?
Lowl3v3l has joined #nixos
romildo has joined #nixos
__Sander__ has quit [(Quit: Konversation terminated!)]
<ixxie>
(leaving out for the moment the complication of language specific package managers :P )
<NixOS_GitHub>
[nixpkgs] 7c6f434c pushed 5 new commits to master: https://git.io/vMNqj
<NixOS_GitHub>
nixpkgs/master 0121066 Vincent Laporte: camlidl: fix build with OCaml 4.03...
<NixOS_GitHub>
nixpkgs/master 2f2fbc9 Vincent Laporte: ocamlPackages.mlgmpidl: init at 1.2.4...
<NixOS_GitHub>
nixpkgs/master 67a5833 Vincent Laporte: ocamlPackages.apron: init at 20160125...
georges-duperon has quit [(Ping timeout: 252 seconds)]
<NixOS_GitHub>
[nixpkgs] aherrmann opened pull request #22104: diagrams: Fix the 1.4.* versions of the ecosystem (master...pr_diagrams) https://git.io/vMNO3
<eacamero_>
I want to use security.pam.enableSSHAgentAuth but I've enabled it and I can't figure out how to configure my user and root such that I don't need a password when running sudo.
peti_ has joined #nixos
peti has quit [(Read error: Connection reset by peer)]
<NixOS_GitHub>
[nixpkgs] 7c6f434c pushed 2 new commits to master: https://git.io/vMN35
ixxie has quit [(Remote host closed the connection)]
glines has joined #nixos
avn has joined #nixos
georges-duperon has joined #nixos
derjohn_mob has quit [(Ping timeout: 240 seconds)]
mudri has quit [(Ping timeout: 248 seconds)]
mkoenig_ has joined #nixos
petar has quit [(Remote host closed the connection)]
mkoenig has quit [(Ping timeout: 240 seconds)]
deepfire has quit [(Ping timeout: 256 seconds)]
pxc has joined #nixos
eacamero_ is now known as eacameron
<eacameron>
I want to use security.pam.enableSSHAgentAuth but I've enabled it and I can't figure out how to configure my user and root such that I don't need a password when running sudo.
dbmikus has quit [(Quit: WeeChat 1.7)]
dbmikus has joined #nixos
tvon_ has joined #nixos
nschoe has joined #nixos
tvon has quit [(Ping timeout: 245 seconds)]
<pxc>
so I'm using an old nix install (through a pivot-root) on the computers at my university so I can have some custom packages installed, but I haven't used it in a while so my nix (1.8) is too old to evaluate nixpkgs
<pxc>
the instructions nix spits out don't work
mudri has joined #nixos
johnsonav has quit [(Ping timeout: 255 seconds)]
<pxc>
can I still grab the newest nix binary from hydra and install it? the links on Hydra are all bad now that everything is stored on S3
morsicus1 has joined #nixos
<gchristensen>
pxc: if you find the path of it (like /nix/store/hash-name) you can do nix-store -r <thatpath>
<niksnut>
pxc: you could try something like: nix-store -r /nix/store/qdkzm17csr24snk247a1s0c47ikq5sl6-nix-1.11.6
<niksnut>
that will bypass evaluation
jsgrant- has joined #nixos
morsicus1 has quit [(Client Quit)]
<nschoe>
Hi all, I'm wondering: `nix-garbage-collect` is safe, right? It only deletes paths that are not used by any profiles?
<simpson>
nschoe: nix-collect-garbage is safe. Dunno what nix-garbage-collect is.
<pxc>
niksnut, gchristensen: thanks. running that now
<nschoe>
simpson, yes, sorry I mixed them up.
lsix has joined #nixos
<simpson>
nschoe: No worries, just trying to be precise.
<nschoe>
simpson, they my net question is: is there any reason why this is not run periodically, and automatically by nix? I was going very very low on space, so low in act that I could not update anything anymore, and I ended up deleteing some old generations and run nix-collect-garbage and bah I freed 14GB :-)
<NixOS_GitHub>
[nixpkgs] romildo opened pull request #22106: drake: init at 0.9.2.0.3.1 (master...new.drake) https://git.io/vMNWV
<gchristensen>
nschoe: you can configure nixos to run nix-collect-garbage regularly
<pxc>
niksnut, gchristensen: okay, so after that do I have to do anything additionally to actually install the package? I got a warning that the package might be picked up by garbage collection because I forgot --add-root
johnsonav has joined #nixos
<nschoe>
gchristensen, ah! Any link for this?
<dtzWill>
nschoe: there's some documentation on having it run automatically, actually! it's pretty useful as long as you are okay potentially re-downloading sources and such if you don't keep roots
<NixOS_GitHub>
[nixpkgs] ttuegel pushed 2 new commits to master: https://git.io/vMN4J
<NixOS_GitHub>
nixpkgs/master e38970c Thomas Tuegel: nixos/ibus: fix custom panel example...
<NixOS_GitHub>
nixpkgs/master 54df142 Thomas Tuegel: nixos/kde5: use kimpanel with IBus by default
<NixOS_GitHub>
[nix] domenkozar created osx-minimal-version (+1 new commit): https://git.io/vMN41
<NixOS_GitHub>
nix/osx-minimal-version 9d4889b Domen Kožar: bail out if osx 10.9 or lower is used during installer
<NixOS_GitHub>
[nix] domenkozar opened pull request #1199: Bail out if osx 10.9 or lower is used during installer (master...osx-minimal-version) https://git.io/vMN49
ilja_kuklic has joined #nixos
kier has joined #nixos
kier has quit [(Remote host closed the connection)]
<fresheyeball>
Any idea why this might be happening?
<cutuchiqueno>
yesterday @goibhniu helped me to change a forgotten root password on a machine which I haven't used for a long time and asked me if I could move wiki sec 'Help, I locked myself out!' with my additions to the manual
<swflint>
Corngood: ext4, switched kernel, works better now
<cutuchiqueno>
There is no ticket for this wiki section
<cutuchiqueno>
should I just open one or do it without creating such a ticket?
nschoe has quit [(Quit: Program. Terminated.)]
<domenkozar>
fresheyeball: are you using master?
<fresheyeball>
domenkozar: what do you mean (sorry very new to nix)
<corngood>
swflint: sweet. Better as in how you'd expect for the hardware?
<domenkozar>
fresheyeball: what command are you usin
jluttine has quit [(Read error: No route to host)]
jluttine has joined #nixos
pSub has joined #nixos
pSub has quit [(Changing host)]
pSub has joined #nixos
reinzelmann has quit [(Ping timeout: 245 seconds)]
<swflint>
Okay, I have services.xserver.synaptics.enable set to true, but the touchpad does not work, only the touchscreen.
ambro718 has quit [(Ping timeout: 240 seconds)]
<swflint>
What can I do?
romildo has joined #nixos
tvon has quit [(Quit: System is sleeping...)]
romildo_ has joined #nixos
jgertm has quit [(Quit: WeeChat 1.6)]
Itkovian has joined #nixos
jgertm has joined #nixos
<romildo_>
I am going to submit a PR to update mkvtoolnix, but it depends on PR #22106 (drake: init at 0.9.2.0.3.1), because mkvtoolnix does not bundle drake anymore.
dbmikus has joined #nixos
<gchristensen>
sounds good romildo_, make a note about the dependency in the PR
<avn>
Are zfs boot/root require copying kernels to /efi?
Itkovian has quit [(Client Quit)]
shd has joined #nixos
alx741_ has joined #nixos
<shd>
what's the right way to create /var/lib/* directory for state in nixos module?
Itkovian has joined #nixos
<romildo_>
gchristensen, ok. I was not sure if I would have to wait for the drake PR being merged.
reinzelmann has joined #nixos
<shd>
would like to fix it but i'm not sure whether i should simply mkdir in module, or there's better way?
<swflint>
How can I enable a touchpad and disable a touchscreen?
alx741 has quit [(Ping timeout: 264 seconds)]
<xwvvvvwx>
Hey :)
<xwvvvvwx>
So I'm trying to package polybar
romildo has quit [(Quit: Leaving)]
<xwvvvvwx>
which has wireless tools as an optional dependency
<dtz>
hmm, I'm trying to create a derivation for a swift project (:D)-- current problem is fetching the project and it's dependencies ahead of time... .. Any folks familiar with these things (swift) or how we handle it for other languages that can give me some advice/pointers on doing this sanely? :)
<dtz>
there's a "swift package fetch" thing to fetch deps but I guess I'd have to make my own fetcher thing for that?
<maurer>
dtz: You could look at what buildRustPackage does for cargo stuff
<dtz>
great, I'll take a look thank you! :)
tvon has joined #nixos
reinzelmann has quit [(Quit: Leaving)]
<propumpkin>
pikajude: do you do anything with nix ruby nowadays?
<propumpkin>
LnL: trying to update a ruby package in nixpkgs (fluentd) and can't figure out how to update the gemfile.lock. If I use the nixpkgs bundler, `bundler package --no-install` just complains and fails
rmeis has joined #nixos
<propumpkin>
oh
<propumpkin>
bundler lock seems to do something promising
<LnL>
complains with what?
<propumpkin>
"An error occurred while installing cool.io (1.4.5), and Bundler cannot continue."
<propumpkin>
but `bundler lock` updated the lockfile
<propumpkin>
now bundix is complaining instead :) but I made it farther
<LnL>
by the way, any idea why a build might be failing with -lIOKit
<NixOS_GitHub>
[nixpkgs] shdpl opened pull request #22112: syncthing: create dataDir on PreStart (master...syncthing) https://git.io/vMNyX
avn has quit [(Remote host closed the connection)]
<propumpkin>
LnL: I wouldn't expect -lIOKit to ever work
<propumpkin>
it should be -framework IOKit
rmeis has quit [(Ping timeout: 260 seconds)]
<LnL>
yeah I was wondering that, but I can't find where it's coming from
* LnL
doesn't understand cmake
<propumpkin>
does anyone? :)
<gchristensen>
_can_ anyone? :)
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
jluttine has quit [(Read error: Connection reset by peer)]
jluttine has joined #nixos
<propumpkin>
:)
avn has joined #nixos
rmeis has joined #nixos
SovereignBleak has quit [(Ping timeout: 252 seconds)]
<LnL>
I think I found it, but I get missing symbols now
SovereignBleak__ has joined #nixos
<avn>
`tank/zroot on / type zfs (rw,relatime,xattr,noacl)` it boots ;)
<avn>
also I found interesting layout for mirroring boot /boot/{sda,sdb}/{grub,efi}
<LnL>
hopefully it's a bit more stable now
<avn>
LnL: not sure, but it the way to removing md+lvm from under
<avn>
also I can kill old nixos volume with btrfs, and pass space to tank as well
<swflint>
Yep. Tried libinput too. Should I try both?
athan has quit [(Remote host closed the connection)]
<corngood>
You haven't got any input from the touchpad in any configuration?
<swflint>
Nope.
<swflint>
And they can't both be enabled.
shd has quit [(Ping timeout: 245 seconds)]
<corngood>
Arch wiki pages for libinput, touchscreen, touchpad have some good troubleshooting info. For example, take a look at `/proc/bus/input/devices`
<NixOS_GitHub>
[nixpkgs] peti pushed 9 new commits to master: https://git.io/vMNxc
<NixOS_GitHub>
nixpkgs/master d350650 Peter Simons: hackage-packages.nix: automatic Haskell package set update...
<NixOS_GitHub>
nixpkgs/master 8ea8493 William Casarin: haskellPackages.ReadArgs: remove jailbreak...
<NixOS_GitHub>
nixpkgs/master a8aee18 William Casarin: haskellPackages.barrier: remove jailbreak...
<NixOS_GitHub>
[nixpkgs] rycee pushed 1 new commit to master: https://git.io/vMNxu
<NixOS_GitHub>
nixpkgs/master 4a914f2 Robert Helgesson: overpass: 2.1 -> 3.0.2
<corngood>
Then maybe `libinput-list-devices `
shd has joined #nixos
<corngood>
Just want to make sure it's supported in your kernel first
<swflint>
Well, the device list only shows the touchscreen.
<NixOS_GitHub>
[nixpkgs] dezgeg opened pull request #22117: Aarch64 (ARM64) support (master...aarch64-for-merge) https://git.io/vMNpA
<gchristensen>
!m Dezgeg
<[0__0]>
You're doing good work, Dezgeg!
ilja_kuklic has quit [(Remote host closed the connection)]
<Unode>
Is there any reason why running "su -" would give me a bad terminal (unless I override TERM=xterm): http://dpaste.com/2S0KCSK while "su" works fine? I'm using urxvt and have http://dpaste.com/3H18VEE set on my configuration.nix
<swflint>
Well that's peachy.
<swflint>
Any idea what to do with that, corngood?
<corngood>
swflint: there are other threads with more positive results. It might need synaptics
<swflint>
Okay. What then?
<Unode>
and on a different note, how do I change the resolution of the linux console? I've set boot.loaded.grub.gfxmodeEfi = "1024x768" but it doesn't seem to produce an effect.
<corngood>
Unode: other distros have an 'extended terminfo' package with those missing terminfo files. If you install urxvt to the system it should work
<Unode>
I'm not sure about the "Efi" part of it.
<Unode>
corngood: the thing is, it seems to work fine on my regular user. It only breaks when I "su -".
<corngood>
Did you install urxvt to your user environment?
<Unode>
no, systemwide
<Unode>
i.e. configuration.nix
<corngood>
Hmm, should be able to find the terminfo files then. It's in systemPackages?
<corngood>
swflint: I think you'll need to turn on synaptics. Could you do that, restart x, and post your xorg log? There should be something in there about synaptics devices
Itkovian_ has joined #nixos
Itkovian has quit [(Ping timeout: 260 seconds)]
Itkovian has joined #nixos
bennofs has quit [(Ping timeout: 256 seconds)]
Itkovian_ has quit [(Ping timeout: 276 seconds)]
<NixOS_GitHub>
[nixpkgs] basvandijk opened pull request #22118: ghc-8.0.2: remove unused fetchFilteredPatch function (master...ghc802-cleanup) https://git.io/vMAv1
<corngood>
Unode: I can reproduce the terminfo problem on mine. in the su shell `TERMINFO_DIRS=/root/.nix-profile/share/terminfo:/nix/var/nix/profiles/default/share/terminfo:/run/current-system/sw/share/terminfo`, and `/run/current-system/sw/share/terminfo/r/rxvt-unicode-256color` exists
indi_ has joined #nixos
indi_ has quit [(Remote host closed the connection)]
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
indi_ has joined #nixos
Avaq has joined #nixos
Avaq has quit [(Client Quit)]
filterfish has quit [(Ping timeout: 245 seconds)]
<NixOS_GitHub>
[nix] trofi opened pull request #1201: mk: add support for passing LDFLAGS to libs and bins (master...master) https://git.io/vMAfx
<Unode>
talking about timing... I had this issue for over 2 months and was working around it by setting TERM=xterm on root's rc file
<Guest33668>
Is there anyone that is getting continuous "curl: (35) SSL connect error" when doing imperative installs (e.g. nix-env -f somepath -iA python3.5-lektor)
<Unode>
oh wait what was patched was on another repository. But anyway, the workaround in the comment works for me. Can you confirm corngood ?
<corngood>
Unode: ah, so it's zsh specific
<Unode>
seems like it. I assume you are also using zsh by default
pxc has quit [(Ping timeout: 245 seconds)]
<corngood>
yeah
<Unode>
now if only I could do the same to my fonts issue :)
<Unode>
playing with fontconfig settings doesn't seem to help. Firefox seems to get a few more fonts right so it happens on less websites but it still does happen.
<corngood>
I would try `grep -i` to be safe. maybe post the complete log with just libinput as well?
MP2E has joined #nixos
Aexoden_ has quit [(Quit: No Ping reply in 180 seconds.)]
<corngood>
there's also `config.hardware.trackpoint.enable`, which might be worth trying. also, anything in `lsusb` that sounds touchpad-ish?
<swflint>
Ix.io/1QYd
<shanemikel>
is there something like a "nix shell hook".. that is, a way after the shell environment has loaded, that I can say, now go ahead and source my ~/.nixbashrc ?
<swflint>
Lsusb: ix.io/1QYe
Aexoden has joined #nixos
<corngood>
swflint: that xorg log is with synaptics? everything is using evdev for input
<swflint>
As far as I know, yes.
<shanemikel>
I'm actually overriding NIX_BUILD_SHELL as it is, and pointing it to a script that execs /usr/bin/env bash with --rcfile ~/.nixbashrc ... but I just realized that's causing some problems.. with environment variables
<corngood>
ok, well try that trackpoint option, and post the log using the libinput driver, but I don't see anything so far that indicates the touchpad is being found by the kernel
tvon has quit [(Quit: Peace out, y'all!)]
<shanemikel>
the problem is: I'd like to clear my environment *before* nix sets it up, and then source my ~/.nixbashrc
<swflint>
The trackpoint setting didn't help.
<Unode>
for the thunderbird+enigmail users in the room. Do you have any problem signing/(de)encrypting messages? Do you see a pinentry window if a message needs to be decrypted or an immediate failure? If you open the keychain externally can you see the decrypted message?
<corngood>
shanemikel: there's a shell hook, for `nix-shell`, I think
<swflint>
Ix.io/1QYh and ix.io/1QYi
georges-duperon has joined #nixos
asymmetric has quit [(Ping timeout: 240 seconds)]
romildo_ has quit [(Quit: Leaving)]
shanemikel has quit [(Quit: My MacBook Air has gone to sleep. ZZZzzz…)]
mudri has joined #nixos
djvdorp has quit [(Ping timeout: 252 seconds)]
<shachaf>
It looks like Nix is having trouble building a package on CentOS 6 without support for user namespaces.