Supersonic112 has quit [(Disconnected by services)]
Supersonic112_ has joined #nixos
Supersonic112_ is now known as Supersonic112
moet has joined #nixos
rprije has joined #nixos
rprije has quit [(Client Quit)]
rprije has joined #nixos
iyzsong has joined #nixos
adisbladis has quit [(Remote host closed the connection)]
adisbladis has joined #nixos
adisbladis has quit [(Changing host)]
adisbladis has joined #nixos
<pmade>
During a package postInstall, the network seems to be unavailable. Is that by design or am I doing something wrong? (read: how do I run something like `npm install` in a postInstall?)
<clever>
pmade: the network is unavailable during the entire build
<clever>
pmade: you must use things like fetchurl to download the files before the build has begun, then copy them into the right spots
<pmade>
clever: That makes sense. I guess I need to figure out how to handle npm install.
<clever>
there is node2nix and npm2nix
<pmade>
clever: Thanks. I'll look into it.
<Ralith>
npm2nix is dead, unless something changed
<Ralith>
also very broken
<clever>
wasnt sure which one had died
eacameron has quit [(Remote host closed the connection)]
<clever>
this is one way to just disable the testing
mkoenig has quit [(Ping timeout: 258 seconds)]
mkoenig has joined #nixos
mbrgm has quit [(Ping timeout: 258 seconds)]
<iqubic>
Also, do I have to install this with nix-build? Can't I put the package in my systemPackages?
<pmade>
When writing package A, Is there a way to refer to an expression that is in a git repo for package B, such that A can depend on B? I'm trying to come up with a way to avoid trying to keep two files in sync.
<clever>
iqubic: just insert this into systemPackages
<catern>
hey #nixos, is there a command which, given some path which is deep within the store (e.g. the absolute path to nix-env), returns the top-level store path for that path? (e.g. /nix/store/...-nix-1.11.10)
sirkha has joined #nixos
moet has joined #nixos
takle has quit [(Remote host closed the connection)]
moet has quit [(Quit: leaving)]
<copumpkin>
catern: cut -d/ -f1-4 ? :)
<catern>
well, I was hoping for something else :)
<catern>
if it handled resolving symlinks and relative paths, that would be mildly convenient :)
<catern>
argh, my dream of doing nix deployment without a tarball is so close yet so far
<iqubic>
Is there a way to speed up my mouse when moving it using my laptop's trackpad? Preferably a solution using services.xserver.synaptics
<catern>
I can copy the Nix closure store paths over to a remote host with an uninitialized /nix/store just fine
<catern>
but then I can't register those paths properly
<catern>
because I lack the appropriate database dump
<catern>
and nix-store --dump-db dumps the entire database...
jushur has joined #nixos
Nobabs27 has joined #nixos
<iqubic>
Like the default mouse speed is too slow for me.
Piece_Maker has joined #nixos
eacameron has joined #nixos
Acou_Bass has quit [(Ping timeout: 248 seconds)]
Piece_Maker is now known as Acou_Bass
eacameron has quit [(Ping timeout: 240 seconds)]
markus1189 has quit [(Ping timeout: 258 seconds)]
markus1189 has joined #nixos
tnias[m] has left #nixos ["User left"]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] ElvishJerricco opened pull request #30072: Fix buildStackProject in nix-build (master...fix-build-stack-project-1.5.1) https://git.io/vdldg
NixOS_GitHub has left #nixos []
Itkovian has joined #nixos
Itkovian has quit [(Client Quit)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] copumpkin pushed 1 new commit to master: https://git.io/vdldb
<NixOS_GitHub>
nixpkgs/master d9091f8 Dan Peebles: xorriso: fix build on Darwin
NixOS_GitHub has left #nixos []
Nobabs27 has quit [(Quit: Leaving)]
rprije has quit [(Ping timeout: 248 seconds)]
<hodapp>
I have libvirtd running via virtualisation.libvirtd.enable; must I run virt-manager as root in order to do anything? or is there some standard way to change permissions?
takle has joined #nixos
<Nobabs27[m]>
whats the correct way to install a binary via nix?
<Nobabs27[m]>
because I got this one, it won't work if I move it out of the folder it comes it...
cpennington has quit [(Remote host closed the connection)]
nh2 has quit [(Ping timeout: 240 seconds)]
<hodapp>
Nobabs27[m]: install a binary via nix? what do you mean?
<Nobabs27[m]>
like a pre-compiled binary
<Nobabs27[m]>
should it be moved to $out/bin or...?
takle has quit [(Ping timeout: 240 seconds)]
fragamus has joined #nixos
iqubic has quit [(Ping timeout: 260 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] peterromfeldhk opened pull request #30073: Ansible/2.4 (master...ansible/2.4) https://git.io/vdlF0
NixOS_GitHub has left #nixos []
justanotheruser has quit [(Ping timeout: 260 seconds)]
hamishmack has quit [(Ping timeout: 240 seconds)]
justanotheruser has joined #nixos
azdle has joined #nixos
sivteck has joined #nixos
sivteck has quit [(Client Quit)]
athan has quit [(Ping timeout: 255 seconds)]
schoppenhauer has quit [(Ping timeout: 258 seconds)]
<hodapp>
ugggh, why does libvirt try to change user + mode of a disk image? it is failing on a CIFS share on which my VM image resides.
schoppenhauer has joined #nixos
justanotheruser has quit [(Quit: WeeChat 1.7.1)]
justanotheruser has joined #nixos
sivteck has joined #nixos
eacameron has joined #nixos
athan has joined #nixos
rprije has joined #nixos
jushur has quit [(Max SendQ exceeded)]
jushur has joined #nixos
shanemikel has joined #nixos
shanemikel has quit [(Client Quit)]
trcm has left #nixos ["WeeChat 1.5"]
samae has quit [(Read error: Connection reset by peer)]
eacamero_ has quit [(Remote host closed the connection)]
jsgrant_ has joined #nixos
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
afics has joined #nixos
eacamero_ has joined #nixos
<jsgrant_>
So, did a 'nixos-rebuild boot --upgrade" and even though I had 17.09 set in config ... moved me to 18.03pre branch -- any clues?
eacamero_ has quit [(Remote host closed the connection)]
<jsgrant_>
Was I supposed to 'nix' the --upgrade if already on that branch?
eacamero_ has joined #nixos
eacameron has quit [(Read error: Connection reset by peer)]
iqubic has joined #nixos
<iqubic>
If I use want to ghcWithPackages to pull hackage deps, where do I put that?
eacameron has joined #nixos
fragamus has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
eacamero_ has quit [(Ping timeout: 260 seconds)]
eacameron has quit [(Remote host closed the connection)]
eacameron has joined #nixos
eacamero_ has joined #nixos
eacamero_ has quit [(Remote host closed the connection)]
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
eacameron has quit [(Remote host closed the connection)]
eacameron has joined #nixos
eacameron has quit [(Remote host closed the connection)]
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
graperoot has joined #nixos
<graperoot>
hellooo
<graperoot>
I'm wondering, should I change the configuration.nix to include the right mount points prior to installation; or does mounting the good paritions on the /mnt/home and /mnt/boot suffice ?
<nixy>
Are you doing a fresh NixOS installation?
<graperoot>
yep =)
<nixy>
Whatever you mount in the disk should be automatically included in your hardware configuration once you run nixos-generate-config then
<graperoot>
alright !
<graperoot>
thanks.
<nixy>
No problemo, just went through a fresh install for 17.09 meself
<graperoot>
it doesn't seem to include the /home and /boot paths..
<graperoot>
mount points I mean
<graperoot>
weird, actually; the nixos-generate-config --root /mnt doesn't do anything anymore. I deleted the config files and tried again. wut wut wut.
<graperoot>
i'll unmount and remount everything and try again.
<graperoot>
yeah, worked. weird though.
<graperoot>
i'm wondering, should I just enable wireless, networkmanager or both ?
sivteck has quit [(Quit: user missing.)]
* jsgrant_
will look into branch-issue tomorrow; Bedtime for now. o/
* jsgrant_
is afk.
<hyper_ch>
graperoot: either wireless or networkmanager... not both
<hyper_ch>
graperoot: I use network manager with kde though
<graperoot>
hyper_ch: thanks
chreekat has quit [(Quit: quitting)]
proteusguy has quit [(Remote host closed the connection)]
katyucha has quit [(Ping timeout: 248 seconds)]
voiceftp has quit [(Ping timeout: 264 seconds)]
<graperoot>
worked fine, time to reboot
graperoot has quit [(Remote host closed the connection)]
katyucha has joined #nixos
voiceftp has joined #nixos
katyucha has quit [(Ping timeout: 248 seconds)]
sivteck has joined #nixos
vandenoever has quit [(Ping timeout: 246 seconds)]
proteusguy has joined #nixos
MercurialAlchemi has joined #nixos
sivteck has quit [(Quit: user missing.)]
JosW has joined #nixos
lsyoyom has quit [(Ping timeout: 246 seconds)]
sivteck has joined #nixos
Cale has quit [(Ping timeout: 258 seconds)]
lsyoyom has joined #nixos
moet has joined #nixos
fragamus has joined #nixos
<MichaelRaskin>
catern: «nix-store -r» normalizes the argument more than you could expect. And prints the result. And if something is already in store, realisation is a no-op (although with some DB reading)
vandenoever has joined #nixos
sivteck has quit [(Quit: user missing.)]
sivteck has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] fpletz pushed 1 new commit to staging: https://git.io/vdlhD
<NixOS_GitHub>
nixpkgs/staging a98b968 Franz Pletz: curl: 7.55.1 -> 7.56.0 for CVE-2017-1000254...
NixOS_GitHub has left #nixos []
reinzelmann has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] fpletz pushed 1 new commit to release-17.09: https://git.io/vdlhH
<NixOS_GitHub>
nixpkgs/release-17.09 a35f8b0 Franz Pletz: curl: 7.55.1 -> 7.56.0 for CVE-2017-1000254...
NixOS_GitHub has left #nixos []
fragamus has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
sivteck has quit [(Client Quit)]
sivteck has joined #nixos
fragamus has joined #nixos
sivteck has quit [(Client Quit)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] fpletz pushed 2 new commits to release-17.03: https://git.io/vdlje
<NixOS_GitHub>
nixpkgs/release-17.03 6158b6b Tim Steinbach: curl: 7.55.0 -> 7.55.1...
<NixOS_GitHub>
nixpkgs/release-17.03 0938f2e Franz Pletz: curl: 7.55.1 -> 7.56.0 for CVE-2017-1000254...
NixOS_GitHub has left #nixos []
jensens has joined #nixos
MichaelRaskin has quit [(Quit: MichaelRaskin)]
moet has quit [(Quit: aeroplane)]
dustmote has joined #nixos
periklis has joined #nixos
Itkovian has joined #nixos
vandenoever has quit [(Ping timeout: 240 seconds)]
reinzelmann has quit [(Quit: Leaving)]
<hyper_ch>
so, slowly getting used to mounting encrypted zfs volumes (already done so twice successfully)
<hyper_ch>
clever: how does ZFS cope with sudden powerloss? I tend to think because of COW it shouldn't do any harm, except that the last changes written may not come in effect.... similar to transactional db queries....
<clever>
hyper_ch: yeah, i believe thats how it works
<clever>
hyper_ch: and at the root level, there is a ring-buffer, pointing to the immutable root directory
reinzelmann has joined #nixos
<hyper_ch>
well, from what I read, file is read from FS, you change it, ZFS writes it back to disk and checksums it... only after this operation the "newly" written file gets assigned as current file
reinzelmann has quit [(Remote host closed the connection)]
<hyper_ch>
at least that's how I understand how cow should work
dustmote has quit [(Quit: bye)]
dustmote has joined #nixos
<sphalerite>
pmade: did you solve your refer-to-A's-git-repository-from-B issue? I'm not sure I understand it correctly but it sounds to me like it should be possible
reinzelmann has joined #nixos
<sphalerite>
hodapp: if you're in the libvirtd group you should be able to use it as yourself too. (Sorry if this is duplicated from what someone else has already said)
dustmote has quit [(Client Quit)]
dustmote has joined #nixos
<sphalerite>
hyper_ch: yes, I think that's right. There's also the ZIL which I don't fully understand but I believe it's the same concept just that it writes it to the ZIL before writing to the "permanent storage" so you have a log that the write was going to happen even if the system lost power before it could write it to the slower disks
<hyper_ch>
what's zil?
<etu>
Write cache afaik
orivej has joined #nixos
<etu>
Many people have a single dedicated ssd for ZIL (I think)
<etu>
Especially if it's spinning rust as storage
dustmote has quit [(Ping timeout: 240 seconds)]
reinzelmann has quit [(Quit: Leaving)]
<clever>
hyper_ch: my understanding, is that when you write to any file, it also has to modify every directory above it in the tree, because the directories are also immutable
<clever>
hyper_ch: until you get to the root of a dataset, and add an entry to the ringbuffer
numerobis has joined #nixos
<hyper_ch>
ok
zzamboni has joined #nixos
zzamboni has quit [(Client Quit)]
zzamboni has joined #nixos
phreedom has quit [(Ping timeout: 240 seconds)]
<sphalerite>
Heh, how about another layer: the HDDs are just a cache for the backing tape-based storage :p
<sphalerite>
Is writing zfs send streams to tapes how tape-based backups are done with zfs?
jushur has quit [(Quit: The Borg joined forces with Skynet, Resistance is futile! Uploading has begun!)]
ThatDocsLady has joined #nixos
<hyper_ch>
Mic92: I know meanwhile :)
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 4 new commits to master: https://git.io/vd8J8
<NixOS_GitHub>
nixpkgs/master d514aa5 David Kuo: Add david50407 into maintainers.nix
<NixOS_GitHub>
nixpkgs/master bc2ac2a David Kuo: crystal: 0.23.0 -> 0.23.1...
<NixOS_GitHub>
nixpkgs/master 766ede0 Orivej Desh: crystal: fix build by using the previous bootstrap
NixOS_GitHub has left #nixos []
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] peterromfeldhk opened pull request #30074: fastlane (master...peter/fastlane) https://git.io/vd8Jg
NixOS_GitHub has left #nixos []
reinzelmann has quit [(Quit: Leaving)]
ertes has quit [(Ping timeout: 264 seconds)]
boxofrox has quit [(Ping timeout: 255 seconds)]
boxofrox has joined #nixos
zraexy has quit [(Ping timeout: 248 seconds)]
katyucha has joined #nixos
goibhniu has joined #nixos
leat has joined #nixos
<fearlessKim[m]>
sphalerite: yep the xps bios looks like a windows phones: tiles of different sizes; you have scrollbars, you can use the mouse, there are hundreds of options
<fearlessKim[m]>
was scared it would crash with so many options
sjourdois has quit [(Quit: ZNC - 1.6.0 - http://znc.in)]
<sphalerite>
fearlessKim: lol
JosW has quit [(Quit: Konversation terminated!)]
<sphalerite>
Mine has lots of options too, and works with the mouse and touchscreen as well as the keyboard
<sphalerite>
But aesthetically it's more windows 95
hamishmack has joined #nixos
vandenoever has quit [(Ping timeout: 246 seconds)]
<fearlessKim[m]>
did I speak about colors ? each page is a rainbow
<fearlessKim[m]>
I admit I haven't checked the touchscreen but I am pretty sure it works
<fearlessKim[m]>
I just read that linus (torvalds) has an XPS too \o/
<Olgierd>
hi
<Olgierd>
I have two luks-encrypted partitions
<Olgierd>
the one (which is /) is protected by a passphrase and the other (HDD) is encrypted with a key file in /etc
<clever>
sphalerite: the bios on one of my machines can even save screenshots to usb
<clever>
sphalerite: another machine, has twinkling stars in the background image
<Olgierd>
how can I configure boot.initrd.luks.* to first open and mount / and then the other partition using the key file
ertes-w has joined #nixos
sjourdois has joined #nixos
<fearlessKim[m]>
it's great that bios don't drop us into imperialist qwerty keyboqrds anymore qwerty anymore
vandenoever has joined #nixos
vandenoever has quit [(Changing host)]
vandenoever has joined #nixos
<fearlessKim[m]>
but I wish they would not create an OS within the OS
phreedom has joined #nixos
<alp>
Olgierd, I've followed http://grahamc.com/blog/nixos-on-dell-9560 -- I think the setup you're describing is slightly different but it might be helpful nonetheless
<sphalerite>
fearlessKim: so does Linus not-Torvalds :p
<fearlessKim[m]>
sphalerite: well you showed him :p
<aw>
Oh, while you're at boot.initrd.luks.*… I just installed 17.09 on a MacBook 1,1 (first generation Intel, 32bit). / (including /boot) is on encrypted btrfs. I configured a keyFile for the device and put it into an extraInitrd and set boot.loader.grub.extraInitrd accordingly.
zzamboni has quit [(Ping timeout: 240 seconds)]
<sphalerite>
fearlessKim: yeah the bios does too much stuff
thc202 has joined #nixos
<nixy>
Is there any sort of stable URL for downloading the NixOS installer ISO
<aw>
When booting, the kernel just panics with "cannot mount root filesystem". If I remove the extraInitrd from grub.cfg it boots fine (of course it fails to open the LUKS device), so it seems GRUB (or the kernel) struggle with the multiple initrds.
<aw>
Any ideas?
Wizek__ has joined #nixos
Wizek has joined #nixos
<aw>
The exact same setup works fine on my ThinkPad X230 (64bit EFI).
<clever>
aw: i have made multiple initrd's work fine with ipxe before
<clever>
aw: to start with, add boot.allow_shell to the kernel args, and get a shell after it fails
<clever>
aw: then try to confirm what files are present after the kernel unpacks everything
<aw>
clever: I wanted to do that, but it seems to load the initrds at all but just dies straight away with a kernel panic because of no-root-fs.
<aw>
:(
<fearlessKim[m]>
gosh the "rebuild buildvm" is so fancy
zzamboni has joined #nixos
<clever>
aw: that would imply it got zero rootfs's, and proceded directly to mounting the rootfs without the aid of an initrd
<clever>
zero initrd's*
<fearlessKim[m]>
wish there was not the dot at the end "The virtual machine can be started by running /nix/store/9z3knfs618s1b7nzhrhqvvcvangzk25l-nixos-vm/bin/run-jedha-vm." because then you double click the VM name and you get the dot :/
<clever>
aw: what does "file" say about the 2nd initrd?, and the 1st?
<aw>
file: command not found. ;)
<clever>
fearlessKim[m]: yeah, that varies between terminals, i try to play it on the same side and add a space or omit it
<sphalerite>
fearlessKim: result/bin/run-jedha-vm
<aw>
Aaaah, I love NixOS. Especially nix-shell…
<fearlessKim[m]>
sphalerite: don't ask me to think. I follow nix orders
<aw>
/nix/store/6qc6...-initrd/initrd: ASCII cpio archive (SVR4 with no CRC)
<clever>
aw: its a bit odd that your main initrd isnt compressed
<fearlessKim[m]>
one can't connect in the VM though. I suppose the hashes in /etc/passwd prevent it ? (machine specific ?)
<fearlessKim[m]>
s/connect/login
<clever>
fearlessKim[m]: i always set users.users.foo.initialHashedpassword
<clever>
fearlessKim[m]: or for tests, initialPassword = "hunter2";
<sphalerite>
fearlessKim: if you ahven't set an initialHashedPassword or authorised pubkeys for SSH, then yeah
<fearlessKim[m]>
I repeat myself but that sounds clever clever
<aw>
clever: Hm, I didn't do anything to have it uncompressed. But maybe the kernel doesn't like intermixed compressed/uncompressed initrds. I'll try an uncompressed extraInitrd.
<aw>
clever: Oh, and I obviously forgot '-H newc' when building the extra-initrd…
<clever>
aw: oh yeah, that also tripped me up when building them by hand on gentoo
<clever>
the default for cpio (only ever used by the kernel) is not supported by the kernel!
<aw>
clever: Yep, that's the problem… works like a charm if you use the correct parameters.
<Dezgeg>
also it's possible to create initrds (even in the proper format) that the cpio tool can extract fine but the kernel can't
<aw>
Sometime's there isn't even anything broken… you're just too stupid. :/
<Dezgeg>
(don't do find -type f | cpio ..., you *do* need to include directories or it doesn't work)
<aw>
clever: Thanks!
<clever>
aw: yep
<clever>
some of the insanity ive done with initrd's in the past
<clever>
i put my wifi drivers, and firmware into the initrd, then rewrote udev's firmware loader in bash
<clever>
and was able to get an ip, and download the rootfs over wifi
ertes-w has quit [(Ping timeout: 255 seconds)]
<Dezgeg>
doesn't the kernel load the firmwares directly from the filesystem nowadays?
<clever>
when i had done it, the kernel ran a special hook program, and passed it an env variable saying what file it wanted
<clever>
udev would then check the fs, give a positive or negative reply by writing to a /sys entry
<clever>
then write the size, and raw contents to a /sys entry
iyzsong has joined #nixos
<Dezgeg>
right, that was in the past I suppose
<Dezgeg>
would be handy if it worked nowadays so you could compress the fw files...
<clever>
i havent investigated how it has changed since then
ertes-w has joined #nixos
<aw>
BTW: You have _no_ idea how hard it is to drag something to boot that's kicking and screaming… NixOS ISO won't boot on MacBook because f*ck you so after many failed attempts with other Linux ISOs (ArchLinux being my favourite but also failing) I booted Ubuntu to install GRUB on hard disk to boot OS X from CD to make it the boot default to be able to load the NixOS USB key manually via this GRUB.
<aw>
Apple builds really neat hardware. m(
<clever>
aw: if you can get any linux kernel to boot, you could try my kexec trick, but it may not like the uefi then
<sphalerite>
aw: gotta love that apple hardware. I was previously on a mbp and I never got the discrete graphics working
<sphalerite>
Not to mention the hoops I had to jump through to get it booting as well
fragamus has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
vandenoever has quit [(Ping timeout: 246 seconds)]
<aw>
sphalerite: If I wasn't a coward, I'd try CoreBoot on this MacBook. But you need to flash it through an SPI flasher and I rather like it working at all…
<sphalerite>
aw: oh, is it an old macbook?
<clever>
aw: i would want to backup the bios via spi first
<aw>
sphalerite: It's a 1,1 (first gen Intel)
<clever>
aw: then you can always recover it via spi as well
<sphalerite>
oh right
<fpletz>
aminechikhaoui: there was an issue which was fixed this morning. please wait for the next channel release and try again
<fpletz>
aminechikhaoui: are you using the systemd user service?
<sphalerite>
yeah mine was more modern than that (2011) and wasn't supported by coreboot
iyzsong has quit [(Ping timeout: 240 seconds)]
<aw>
clever: Yeah, but you'll have to open it up until you get down to the flash and even then I'd be much to worried to inadvertently fry the mainboard…
<aw>
On the other hand: in the worst case it's a MacBook… you could still claim that it wasn't inadvertently. :P
<sirkha>
Anyone successfully have qt4 apps using GTK style?
<sirkha>
I did a package override to enable the GTK style plugin, then did nix-env -i qpdfview, but nothing changed
<fpletz>
aw: that's very difficult if you don't apply external power :) I've flashed lots of thinkpads with coreboot recently
<sirkha>
I used nix-store to find the connected QT dependency, and it, as far as I can tell, did not build the gtkStyle plugin
<aminechikhaoui>
fpletz: ok awesome will probably point to a local checkout of the channel, which systemd user service ? I'm just using the 2 options programs.gnupg.agent.enable and enableSSHSupport
<sirkha>
even if it had, nothing seems to add the qt4 plugins in that store to the user environment
mudri has quit [(Ping timeout: 240 seconds)]
<fpletz>
aminechikhaoui: that's the one :) you can have a look at the logs with: journalctl --user -f -u gpg-agent
<aw>
fpletz: My other notebook (the »real« one, though) is a ThinkPad X230. This should be supported, too, but same thing… CoreBoot wobsite claims you have to be really careful as the flash seems to be very fragile, so no CoreBoot for the coward.
<aminechikhaoui>
fpletz: oh ok thanks, yeah was trying to look at the logs with journalctl -f gpg-agent and missed that it's a systemd.user.* service
<fpletz>
aw: most of them were x230s… I've yet to fry one and I had to apply external power because I was initially flashing with a buspirate :)
<fpletz>
aw: I can recommend the ch341a-based flashers, they're very fast and supply enough power
<aw>
Hm, I'll make a note about that. But seriously, I doubt I'll build up the courage before I can afford a new one. But then I'll do, I'm definitely looking forward to this. :)
__Sander__ has joined #nixos
<fpletz>
aw: if you're at nixcon or in munich sometime I can flash it for you
<aw>
In the meantime I'm considering buying a used X200t or so. Those are supported as well an becoming insanely cheap lately. And I always wanted to play with a tablet one. ;)
nwspk has quit [(Ping timeout: 240 seconds)]
<aw>
fpletz: I'd really love to. I'm currently in Dresden, thus it wouldn't even be too far, but I'm kind of bound here. My daughter just got nine weeks old. :)
lsix has quit [(Quit: Konversation terminated!)]
<fpletz>
aw: oh, congratulatios :)
<aw>
fpletz: Thank you. :)
lsix has joined #nixos
<aw>
fpletz: I think I should let _her_ do the flashing once she's old enough. :D
Acou_Bass has quit [(Ping timeout: 248 seconds)]
<aw>
fpletz: »And this is how you build a paperweight from used parts.«
justbeingglad has joined #nixos
Acou_Bass has joined #nixos
nwspk has joined #nixos
xd1le has joined #nixos
<sphalerite>
haha
<sphalerite>
I really want a coreboot device
<sphalerite>
I wish my laptop supported coreboot. I could probably make it boot so much faster. The stock BIOS takes like 10s to hand over to systemd-boot :(
<makefu>
sphalerite: you can get an x220 or x230. unfortunately your battery will last shorter because power management does not seem to be "core" enough to be integrated
<sphalerite>
and it's also not as powerful as my main laptop
justbeingglad has left #nixos []
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] woffs opened pull request #30076: unicode: change UnicodeData.txt to stable version (master...unicode.txt) https://git.io/vd8mc
NixOS_GitHub has left #nixos []
<sphalerite>
I have an Asus C201 actually, have yet to install my own coreboot on it though. I thought I'd try to get the stock firmware booting my own kernel first though
<sphalerite>
idk if that's actually a useful endeavour though…
<sphalerite>
I'm also just scared to brick it, because I don't have a hardware flasher thingy
sigmundv_ has joined #nixos
<fpletz>
sphalerite: a spi flasher plus a soic clip and some cables should cost under 50eur :)
<fpletz>
and a used x230 in good condition should cost around 300eur
<sphalerite>
Is there a reason to go for an x230 over just using the chromebook I already have?
<sphalerite>
actually, I wonder how the CPUs compare power-wise
<fpletz>
oh, I didn't realize that's a chromebook
<ij>
Can I install nixos-generate-config and all the tools from installer/tools/tools.nix with nix on another distribution?
<fpletz>
I need the trackpoint to get stuff done, I can't work with touchpads anymore :)
<sphalerite>
fpletz: and it's an ARM chromebook
* fpletz
even has an external thinkpad usb keyboard with trackpoint :)
<sphalerite>
so it's already running coreboot
<sphalerite>
I just want my own coreboot with my own payload because… I don't even know why
<sphalerite>
because it's cool :D
<fpletz>
well, makes a lot of sense to me :)
<sphalerite>
and I can flash it purely with software. I just need the flasher if I brick it in that process
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 2 new commits to master: https://git.io/vd8YI
<NixOS_GitHub>
nixpkgs/master 12f9b43 Frank Doepper: unicode: change UnicodeData.txt to stable version...
<fpletz>
I use grub as payload instead of seabios, grub can unlock my luks partition to get the grub.cfg, kernel and initrd… directly from the rom and I have real full disk encryption :)
<fpletz>
another thing I haven't finished: secure boot with grub :) grub can verify gpg signatures, so you can configure it to only load kernels and initrds you've signed with your gpg key :>
<xd1le>
when installing NixOS from a live usb, if I make files under my user home directory (so /mnt/home/user), because I'm installing from the root account, are these files going to have root permissions even after installation? or will the nixos-install script fix the permissions (if I specify that /home/user is the home directory for my user)?
<sphalerite>
Not sure that'll fit on a regular consumer-grade flash chip though
fendor has joined #nixos
<sphalerite>
and is there really any point in encrypting the bootloader and kernel and stuff? I don't see why you'd need to keep that secret. Of course integrity is important, but that's what signatures (as you mentioned) are for
proteusguy has quit [(Remote host closed the connection)]
<sphalerite>
hm, not sure I could actually use grub for my chromebook. I'd have to have OVMF load GRUB AFAIU
<fpletz>
sphalerite: yeah, I agree
<fpletz>
cool, never head of petitboot
<sphalerite>
at that point I might as well use OVMF to load linux, or u-boot
<sphalerite>
Dezgeg: nice, didn't know that (and the wikipedia article doesn't mention it :p)
maiksen has joined #nixos
kier has quit [(Ping timeout: 240 seconds)]
<hyper_ch>
hmmm, when I'm in a 17.03 install usb environemnt, how can I setup a new system that directly uses nixos unstable channel?
Ridout has joined #nixos
<maiksen>
Hey, this may seems like a silly question, but should installing nixos work on mbr/gpt hybrid environments? My old gen8 proliant homeserver does not support uefi (at least as far as I know) and so far I failed installing nixos on that system.
<etu>
maiksen: if you don't use uefi you probably use mbr?
<aminechikhaoui>
fpletz: so I'm now seeing this when using pass to get my passwords in the gpg-agent journal "command 'PKDECRYPT' failed: No such file or directory <Pinentry>"
<aminechikhaoui>
any idea if it's related to any recent commit ?
proteusguy has quit [(Remote host closed the connection)]
<maiksen>
etu: thats what I thought, but I read that I need to do that bios boot partition "workaround" as my disk is bigger than 2tb
<sirkha>
can still use the gpt partition tables with that type of bios?
mbld has joined #nixos
<maiksen>
but from what you wrote I guess it should work so the problem seems to be me :) not some silly bug I dont know about
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Mic92 pushed 1 new commit to master: https://git.io/vd8Cq
<NixOS_GitHub>
nixpkgs/master a320034 Joerg Thalheim: nixos/traefik: owner/group should be changed recursivly
NixOS_GitHub has left #nixos []
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] domenkozar pushed 1 new commit to master: https://git.io/vd8C3
<NixOS_GitHub>
nixpkgs/master 2b70f5e Domen Kožar: jq: sanity check with --help
NixOS_GitHub has left #nixos []
<maiksen>
etu: thanks, fpletz thanks :)
<hyper_ch>
clever: for zfs install, you need a seperate boot partition right? you can't have /boot in the zfs pool?
<fpletz>
aminechikhaoui: hrm, I haven't seen that error before :( that commit will probably not fix that
<aminechikhaoui>
also if it might be relevant "nixos pinentry[27261]: Remote error from secret service: org.freedesktop.DBus.Error.ServiceUnknown: The name org.freedesktop.secrets was not provided by any .service files"
<aminechikhaoui>
that's also in the logs
katyucha has quit [(Ping timeout: 248 seconds)]
<fpletz>
ah, then it's probably related… the problem is that the default was changed to the gnome3 version of pinentry and that needs weird dbus services that are probably only available if you're running gnome
<fpletz>
the commit reverts it back to the simple gtk2 version which should work
<fpletz>
aminechikhaoui: what windowmanager or desktop environment are you running?
<aminechikhaoui>
fpletz: plasma5
<fpletz>
ok, then that commit should indeed fix your issue :)
<aminechikhaoui>
hm weird, so I did nixos-rebuild switch -I nixpkgs=channel:nixos-17.09-small and that's after that that I had the issue
<hyper_ch>
Mic92: if you're here, I got question :)
<aminechikhaoui>
afaik the commit was already propagated to the nixos-17.09-small channel
<fpletz>
have you logged out and back in?
<aminechikhaoui>
nope
<fpletz>
the systemd user sessions are unfortunately not reloaded
<aminechikhaoui>
oh ok, let me try that
<fpletz>
if there's still a session running it might even need a reboot
zzamboni has joined #nixos
<aminechikhaoui>
fpletz: phew, now it's working after the reboot
<aminechikhaoui>
\O/
<maiksen>
fpletz: should the layout be like: 1. partition: bios boot partition (no fs), 2. partition: /boot ef01 mbr partition as vfat, 3. partition: root for nixos - and as boot.loader.grub.device I specify just the disk these partitions are on?
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] WilliButz opened pull request #30077: nixos/tests: clean up pgjwt test (master...fix-pgjwt-test) https://git.io/vd8W3
NixOS_GitHub has left #nixos []
<fpletz>
maiksen: sounds correct :) as you're using grub the /boot partition can also be ext2 or whatever you prefer
<fpletz>
aminechikhaoui: cool :)
<maiksen>
okay thank you very much I'll give it another shot
<hyper_ch>
Mic92: can get zfs unstable into live usb environment
<__monty__>
I'm still getting to grips with using nix for haskell development. I have dependencies somewhat figured out, main issue now is getting my editor to play nice.
<__monty__>
I use vim with neomake which in turn uses ghc-mod to check for errors and for linting.
<__monty__>
Outside of a nix-shell neomake complains about a missing cabal and installs a cabal-helper in ~/.ghc-mod; inside of a nix-shell :GhcModCheck & Lint work fine but :Neomake seems to hang.
<hyper_ch>
Mic92: hmmm, it also wanted a hostid from me....
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] joachifm pushed 2 new commits to master: https://git.io/vd84e
<NixOS_GitHub>
nixpkgs/master 3539e16 WilliButz: nixos/tests: clean up pgjwt test...
<NixOS_GitHub>
nixpkgs/master 8021c8f Joachim F: Merge pull request #30077 from WilliButz/fix-pgjwt-test...
NixOS_GitHub has left #nixos []
reinzelmann has quit [(Quit: Leaving)]
iyzsong has quit [(Ping timeout: 255 seconds)]
<clever>
hyper_ch: grub has trouble reading zfs, so its best to have a seperate partition for boot
iyzsong has joined #nixos
<hyper_ch>
clever: ok
<hyper_ch>
also, zfs root wanted a hostid for the system
<clever>
yeah
<hyper_ch>
how does one generate a hostid?
<clever>
the docs for the hostid tell you
<clever>
read the configuration.nix man page
<hyper_ch>
so much work :)
bin7me has joined #nixos
mb[m]2 has left #nixos ["User left"]
iqubic has quit [(Ping timeout: 260 seconds)]
<clever>
"man configuration.nix\n/hostId\n"
<clever>
hyper_ch: how is typing the above "so much work" ?
<hyper_ch>
it's not so much work ;)
<domenkozar>
the question itself had more chars
<hyper_ch>
btw, muscle memory :) it just gets autotyped
<hyper_ch>
(and no, "so much work" is not my encryption password)
dustmote has joined #nixos
Ridout has quit [(Quit: Lost terminal)]
zzamboni has quit [(Ping timeout: 258 seconds)]
dustmote has quit [(Client Quit)]
mbld1 has joined #nixos
mbld has quit [(Ping timeout: 246 seconds)]
acarrico has joined #nixos
<sirkha>
I put qpdfview = pkgs.qpdfview.override { qt4 = pkgs.qt48full; }; in packageOverrides in my config.nix, but it gives me the error, "attribute qt48full' missing
<kuznero>
Is it enough to have kubernetes master installed on local developer station with `services.kubernetes = { roles = [ "master" "node" ]; };` ?
<sirkha>
I am guessing because i am not passing in the full pkgs, but how do i do this correctly?
<kuznero>
disasm: maybe you can help?
<disasm>
kuznero: you would need etcd as well
<kuznero>
disasm: I've also tried this (inspired by your devops-kubernetes repo): https://pastebin.com/PyQ9FMg0 Does that look approximately right?
<disasm>
actually, etcd should be setup by default on the master
<disasm>
err, nm, that just is the config pointing at etcd
<disasm>
I haven't tested running kubernetes on a single node, but that looks like it should work.
freusque has quit [(Ping timeout: 240 seconds)]
guillaum1 has quit [(Ping timeout: 248 seconds)]
<hyper_ch>
Mic92: clever: I managed to get an encrypted zfs root install up and running (in a vm)
<chakerbenhamed>
Hey, is there someone from NixCon organizers online right now?
<disasm>
kuznero: is that 17.09?
<infinisil>
Ah nice, the website now links to 17.09 :D
proteusguy has joined #nixos
<kuznero>
disasm: yes
<hyper_ch>
infinisil: 17.09 iso can be downloaded
katyucha has joined #nixos
<hyper_ch>
needed to that to get proper zfs :)
<hodapp>
sphalerite: oh, good to know, thanks (re: permissions & users in libvirtd group)
<infinisil>
you can also just rebuild when having booted the iso, or create a custom iso with zfs support already
<disasm>
kuznero: by first glance seems like something is trying use network during build.
<hyper_ch>
infinisil: I did use 17.03 before, added the zfs thingy and zfs unstable thingy for encryption but then it didn't wokr... only worked when I used 17.09 iso
<infinisil>
hmm i see, can't think of why that didn't work though, oh well
<infinisil>
oh, wait, zfsUnstable in 17.03 probably isn't the latest unstable with encryption
fendor has quit [(Ping timeout: 248 seconds)]
<infinisil>
Yeah, zfsUnstable in 17.03 is a 0.7 rc
<kuznero>
disasm: I am behind corporate proxy for which I have environment variables HTTP_PROXY+ setup. It is working for most of other things...
<Mic92>
hyper_ch: I actually wrote this in the wiki
<infinisil>
am jeaulous of your zfs encryption
<hodapp>
you should be
<hyper_ch>
btw, I'm pondering to a dataset for VMs.... so that they don't get snapshotted as often as normal stuff
<infinisil>
i'm still living in the stone age with luks
noam has quit [(Read error: No route to host)]
<hyper_ch>
and use like /VMs
<hodapp>
hyper_ch: zvols can be handy there too
<hyper_ch>
infinisil: well, only tested enc. zfs in a vm this far
ShalokShalom has quit [(Ping timeout: 248 seconds)]
<infinisil>
ah i see
<hyper_ch>
hodapp: you're using again one of those big words that mere mortal human beings like me don't know :)
noam has joined #nixos
<infinisil>
hodapp: Why are zvols better for VMs?
<hyper_ch>
Mic92: the wiki is good.. I might have overlooked that about 17.09
<hyper_ch>
Mic92: once you have a dataset, you still can add another device as mirror to it, right?
<hodapp>
Alright, so now I can at least run virsh & virt-manager as a normal user, but I still have the minor annoyance that I can't put disk images on a CIFS share because it insists on changing owner & mode, which fails (though I added noperm), and I haven't tried again yet to see if I can put images in my homedir
<disasm>
kuznero: still waiting on my nodes to finish infecting...
<hyper_ch>
hodapp: why use zvol for VMs?
<Mic92>
hyper_ch: I think so
<hodapp>
depends on what sort of VMs you're using, but it lets you piggyback off of the snapshotting and compression of ZFS while still presenting a block device, as I understand it
numerobis has quit [(Read error: Connection reset by peer)]
<infinisil>
oh and because VM's usually have their own filesystems, it's probably faster
<hodapp>
it might be
<infinisil>
because otherwise it would be a filesystem in a filesystem
<hodapp>
though I'm poking around and seeing that qcow2 still has some pretty nice features
<hodapp>
which I'm now trying to learn
<hyper_ch>
Mic92: that's what I've been doing this far with mdadm... made raid1 with 1 disk... then when I feel like it, I attach a usb 3 drive, extend the raid1 to 2 disks, and tell it to mirror to the usb 3 device... that way I created a hotcopy from which I quickly can boot up again if needed
<hyper_ch>
hodapp: uisng VBox :)
<hodapp>
I am just about to kick vbox to the curb
<hodapp>
after disasm had mercy on one of my incessant 1 AM profanity-filled rants and told me to just use something else
<Mic92>
hyper_ch: `zfs send` sounds more practical for that use case
<infinisil>
hodapp: what was the problem?
<hyper_ch>
Mic92: probably :)
<hyper_ch>
Mic92: the reason was, with the extern usb also in raid1... I could just boot from there... well, I had to fix first the raid obviously but then I could just boot off it
<hyper_ch>
then rsync back the latest backup that I made and could work
<hodapp>
infinisil: needed RDP support. couldn't install the Oracle extension pack. couldn't easily find the docs for how to do that in NixOS. had to completely rebuild VirtualBox to get that, which took annoyingly long. still did the configuration wrong and had to rebuild yet again to get the kernel modules right.
<xd1le>
omg
<xd1le>
vim is in the nixos install image
adc has joined #nixos
<xd1le>
since when
<xd1le>
yees :)
<hodapp>
infinisil: and I've had stuff like this happen on pretty much every platform where I've run VirtualBox over the years.
oida has joined #nixos
<hodapp>
whatever use case was just proposed, yes, probably use 'zfs send'
<hodapp>
I use rsync all the time, but if you're dealing with big filesystems, snapshotting & send/recv can do an incremental update in a fraction of the time it will take
<hyper_ch>
ok, I get it... if in doubt, use zfs send
<hodapp>
sure, if your needs involve keeping two ZFS filesystems
oida_ has quit [(Ping timeout: 258 seconds)]
<infinisil>
hodapp: Oh that sucks, I'll be keeping an eye out if I ever need to use VBox for something for intricate
<hodapp>
infinisil: this wasn't even intricate stuff; most of the time I just have 2-3 VMs that need to run reliably
infinisil has quit [(Quit: gotta go!)]
<hodapp>
so I'm learning virsh & virt-manager now. I still don't entirely understand the flow of things, but I guess now it's virsh/virt-manager -> libvirt -> QEMU/KVM
griff__ has joined #nixos
griff_ has quit [(Ping timeout: 248 seconds)]
griff__ is now known as griff_
<kuznero>
disasm: does that make sense to open a ticket on this?
fendor has joined #nixos
lsix has quit [(Quit: Konversation terminated!)]
<disasm>
kuznero: my droplets just finished infecting, I'll let you know in 5 mins if I can reproduce
<kuznero>
disasm: ok, thanks
<hodapp>
this would sound rather strange out of context
<hodapp>
I guess I was just so used to seeing the shell upon failure from every time my Arch install was screwed up that I figured it was default behavior
<hodapp>
...that, I do not miss
MercurialAlchemi has joined #nixos
<ertes-w>
disasm: that's not the point… i need to do something *new* and debug it when it fails
<ertes-w>
symphori1: thanks
<ertes-w>
my configuration.nix manpage doesn't mention that option though
<symphori1>
that's a normal kernel cmdline parameter
<ertes-w>
ah, i see
peacememories has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
iyzsong has quit [(Ping timeout: 248 seconds)]
cpennington has joined #nixos
pie_ has quit [(Ping timeout: 240 seconds)]
orivej has quit [(Ping timeout: 240 seconds)]
<disasm>
kuznero: works here with my 3 node setup
iyzsong has joined #nixos
<kuznero>
disasm: also with 17.09 behind proxy?
taaperotassu has quit [(Ping timeout: 258 seconds)]
orivej has joined #nixos
<disasm>
no proxy here, but with 17.09
<disasm>
this is what looks odd to me: Error getting v2 registry: Get https://gcr.io/v2/: proxyconnect tcp: dial tcp 127.0.0.1:3128: getsockopt: connection refused
<disasm>
I'm not sure why it's trying to connect to that external registry. Is that defined somewhere on the host?
<kuznero>
disasm: I am running cntlm proxy locally which might have some limitations when fetching through https perhaps?
<disasm>
I have sandboxing enabled, so if mine tried to do that it should have failed as well
<kuznero>
nope, that is not explicitly defined anywhere
<kuznero>
hm...
<disasm>
kuznero: is this a new host or is there existing things on it?
<kuznero>
it is my developer station - so, there are a log of things on it already
Arcaelyx has joined #nixos
mbld1 has joined #nixos
simukis has quit [(Quit: simukis)]
simukis has joined #nixos
<kuznero>
disasm: need to run now (sorry). Thanks for help! If you have some thoughts please send anyway - I will get back to it on botbot.me later on. Thanks again!
<xd1le>
makefu: thanks. but isn't it nixos-config=/path/to/file.conf ?
<makefu>
wupps, yes true
<makefu>
i read configuration, wrote it but meant nixos-config
<xd1le>
makefu: cool, thanks!
<xd1le>
makefu: and I'm assuming this also works for settings nixpkgs, i.e -I nixpkgs=/path/to/nixpkgs
<xd1le>
?
<makefu>
yes exactly
<xd1le>
makefu: cool
kus_ubuntui686 has joined #nixos
kus_ubuntui686 has quit [(Max SendQ exceeded)]
<Mic92>
adisbladis: we do backports on our own usually using: "git cherry-pick -x", so it is enough if you request a backport in pull request description.
<adisbladis>
Mic92: Check :>
jb55 has joined #nixos
hyper_webirc has joined #nixos
<hyper_webirc>
clever: Mic92 : infinisil : so, making dd backup now before wiping the ssd in preparation for encrypted zfs ;)
<cement>
well, that's a fun one. I've been trying to move my vim config to a nix file, and the package is named myVim, but I'd used nix-env to install vim at some point, so using vim as the command didn't work properly
<hyper_webirc>
(need usb 3.1....)
<adisbladis>
hyper_webirc: Tell me how it goes :) I'm planning to migrate my laptop to that this weekend
<hyper_webirc>
adisbladis: to encrypted zfs root?
<adisbladis>
Yes
zzamboni has joined #nixos
<cement>
I nix-env -e'd vim, and then rebuilt, but I still can't use vim as the command
<hyper_webirc>
works fine... tested it first in a vm... and Mic92 did write a nice wiki entry
<hyper_webirc>
but you need 17.09 iso
<MoreTea>
hyper_webirc, do you need to kill your pool to enable that?
<adisbladis>
hyper_webirc: I'm on master :>
<Mic92>
MoreTea: it is just a zpool feature
<hyper_webirc>
MoreTea: I don't have zfs this far... curently it's mdadm raid 1 -> luks/dm-crypt -> ext4
<hyper_webirc>
MoreTea: you just need zfs from unstable for encryption
<MoreTea>
Ah, I see. Can one encrypt an existing zfs fs?
<Mic92>
you can use `zfs send` for migration in the same pool
<hyper_webirc>
MoreTea: you create a new encrypted dataset
<MoreTea>
That makes sense.
<hyper_webirc>
thx Mic92 for the wiki :)
<hyper_webirc>
Mic92: I still tend to think that encfs or something is better as dataset instead of "root" :)
<cement>
ok, nvm
<xd1le>
if I had a GRUB installed from a previous NixOS installation (which has now been wiped), do I need to set boot.loader.grub.device, or would "nodev" still work?
<clever>
xd1le: i would still set boot.loader.grub.device
katyucha has quit [(Ping timeout: 248 seconds)]
<xd1le>
clever: does something like "/dev/sda" sound okay?
<xd1le>
clever: so there's no harm in reinstalling grub?
<clever>
correct
<xd1le>
thanks
<MoreTea>
hmm, this needs a separate boot partition.
<hyper_webirc>
MoreTea: yeah, according to all the zfs wizards in here, nixos has issues with booting from zfs
<hyper_webirc>
(or something like that=
<MoreTea>
well no, I'm booting directly from ZFS with grub
<clever>
grub has trouble traversing really large directories like /nix/store/
<clever>
and if /boot is on the same filesystem as /nix/store, nixos will just use symlinks by default
<MoreTea>
But I can imagine that encrypted ZFS doesn't work yet properly with GRUB
<MoreTea>
ah, they're different datasets on my machine ;)
<hyper_webirc>
MoreTea: with seperate boot it works just fine... tested it in vm
<clever>
MoreTea: there is also a copyKernels option, that forces nixos to copy the kernel, within the same dataset
<adisbladis>
MoreTea: I will usually put grub on a separate ext4 partition
<adisbladis>
Regardless of the rootfs
<clever>
yeah, i just always ext4 it
<MoreTea>
actually /boot is on my root dataset, /nix is on a different dataset.
katyucha has joined #nixos
<alp>
hi folks. looks like my touchpad supports multitouch with just libinput (e.g I can do two finger scrolling), but the touch screen seems to only respond to one finger at a time. does that mean I must enable services.xserver.multitouch? the documentation is a bit scarce, so if anyone knows, that'd be appreciated =)
<hyper_webirc>
even the zfs wizards have some <3 left for ext4 :)
<adisbladis>
hyper_webirc: You don't really see the downsides of ext4 if it only lives in /boot ;)
chakerbenhamed has quit [(Quit: WeeChat 0.4.2)]
griff_ has joined #nixos
<hyper_webirc>
adisbladis: I haven't seen the downsides of ext4 this far even if it lives as /
zraexy has joined #nixos
<clever>
hyper_webirc: i have seen a few new users improperly shutdown nixos after a nixos-rebuild, and it truncated various files in /nix/store/
<hyper_webirc>
clever: then they can just reboot in an old generation :)
<clever>
even if they boot an old generation, nix will get upset that files in the store are corrupt
<clever>
and it wont repair them automatically
<hyper_webirc>
:)
<hyper_webirc>
and this can't happen with zfs?
<hyper_webirc>
I'm pretty sure I can make that happen with zfs
<clever>
zfs journals both the data and metadata, so the files cant get out of sync like that
<clever>
zfs also has checksums on everything, so corruption wont get to the userland apps
<hyper_webirc>
well, I have good experience at trashing btrfs which is supposed to have all those cool features as well
bin7me has quit [(Remote host closed the connection)]
justbeingglad has joined #nixos
Itkovian has quit [(Ping timeout: 240 seconds)]
<etu>
hyper_webirc: But btrfs isn't good :p
justbeingglad has left #nixos []
<hyper_webirc>
clever: I noticed
Sonarpulse has joined #nixos
<etu>
hyper_webirc: I have plenty of experience with corrupt btrfs, without even trying.
vandenoever has quit [(Ping timeout: 260 seconds)]
markus1189 has quit [(Ping timeout: 258 seconds)]
markus1189 has joined #nixos
dieggsy has quit [(Ping timeout: 246 seconds)]
<adisbladis>
Btrfs sadly never matured enough.. It was not that long ago I had data corruption bugs with btrfs send/recv
<hyper_webirc>
I was able to corrupt btrfs without even using those advanced features
csingley has joined #nixos
<adisbladis>
hyper_webirc: zfs saved my bacon on multiple occasions
ThatDocsLady has quit [(Quit: Leaving)]
<adisbladis>
It even told me about broken hardware I had no idea was broken :>
<adisbladis>
Once a loose cable
<__monty__>
I'm assuming this is on servers with plenty of disks?
<adisbladis>
__monty__: Generally yes, also got early warning signs of disk breakage on the laptop
<hyper_webirc>
still another 600GB to dd
<hyper_webirc>
well, I'll add another dataset for VMs.... it seems a good idea
<adisbladis>
It's free anyway ^_^
<hyper_webirc>
and it seems a lot of people have a <3 for a home dataset
takle has quit [(Remote host closed the connection)]
danyc0 has quit [(Ping timeout: 252 seconds)]
<hyper_webirc>
still undecided if I should make a home dataset
<tg>
where's the documentation about what <tags> are availabe in description fields?
<hyper_webirc>
raul: in the configuration.nix, just add the entry for androidsdk.... it should pull all the rest
graperoot has joined #nixos
<graperoot>
hellooo
mbld has joined #nixos
<graperoot>
How do I install deluge on nixos ? nix-env -i deluge yields "matches no derivations"
<raulraja>
@hyper_webirc I install android-studio in the unstable channel that is supposed to do that without modifying my config right? Are there any docs that show what I need in my config if not?
cpennington has quit [(Ping timeout: 248 seconds)]
<makefu>
graperoot: for me it is nix-shell -p deluge
<hyper_webirc>
raulraja: not sure if android-studio puls androidsdk
nh2 has joined #nixos
<hyper_webirc>
nh2: fiber7 :)
<adisbladis>
hyper_webirc: Android-studio likes to manage android sdks by itself
<hyper_webirc>
adisbladis: yeah, no idea :) he mentioned android-studio later... never used it... just need android tool for adb
<xd1le>
hmm so I did a nixos-install and got a no space left on device error, even though the target filesystem has 200G free. the target file system is mounted on /mnt, but I see that there's no /mnt/nix directory
<hyper_webirc>
xd1le: df -h
<xd1le>
my /nix/store is 100% full after checking with df -h
mortum5 has quit [(Ping timeout: 240 seconds)]
<xd1le>
2.2G
<xd1le>
used
<raulraja>
ok, thanks I'll try adding the android-sdk manually
<hyper_webirc>
xd1le: did you run nixos-generate-config --root /mnt ?
<xd1le>
I thought that nixos-install would install to /mnt
<hyper_webirc>
or is /mnt/nixos the root?
<xd1le>
yes
<xd1le>
there's no /mnt/nixos
cfricke has joined #nixos
<hyper_webirc>
do you have config files in /mnt/etc/nixos/ ?
<xd1le>
should I pass --root /mnt option to nixos-install even though it says that it does that by default?
<xd1le>
I do but I'm using -I to override it
<hyper_webirc>
well, try to give the --root param with the install
<xd1le>
just tried and same error
<xd1le>
:/
<adisbladis>
hyper_webirc: You dont happen to know anything about using smartcards with zfs encryption?
<hyper_webirc>
adisbladis: I started testing zfs like last friday ;)
<adisbladis>
Ahh :P
mbld has quit [(Quit: Leaving.)]
<hyper_webirc>
xd1le: always worked for me....
Itkovian has joined #nixos
<hyper_webirc>
adisbladis: Mic92 is the encrypted root guru here
<hyper_webirc>
adisbladis: Mic92 is the encrypted root gzfs uru here
<adisbladis>
Because I have been using my yubikey with luks before and I'd like to keep doing so if possible
<mguex>
Any pypi2nix users here? What is the equivalent for PYTHONPATH : ${python.sitePackages} in wrapProgram, or is this even possible?
<xd1le>
maybe then my first install should be minimal.. sigh
<clever>
xd1le: can you gist the full error and df output?
<xd1le>
should I clean /nix/store with nix-collect-garbage -d
jb55 has quit [(Ping timeout: 248 seconds)]
<xd1le>
clever: okay
<graperoot>
doesn't nix-shell -p only test the package ?
<graperoot>
makefu
<hyper_webirc>
adisbladis: smartcards are in general a pain IMHO
<Mic92>
adisbladis: you can use a key file instead of password for decrypting, if your yubikey can derive such a key/decrypt, then you are good to go.
infinisil has quit [(Ping timeout: 264 seconds)]
raulraja has quit [(Quit: Leaving.)]
infinisil has joined #nixos
<disasm>
graperoot: nix-shell -p starts a shell with that package available
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] TealG opened pull request #30086: axoloti: init at 1.0.12-1 (master...axoloti-1.0.12) https://git.io/vd8bC
NixOS_GitHub has left #nixos []
<disasm>
graperoot: for example, nix-shell -p python27Packages.Fabric will start a shell with python fabric available to be used
<hodapp>
alright, just installed the SPICE guest tools in this Windows VM and everything appears to be working fine with Remmina & QEMU/KVM
* hodapp
nukes VirtualBox from orbit
<disasm>
hodapp: do it! :)
<hodapp>
also, why is this faster? oh well. not complaining.
<graperoot>
disasm : thanks for explaining. can I reuse it later ? Also; Is it possible to make the package available permanently
<graperoot>
?
vandenoever has joined #nixos
tmaekawa has joined #nixos
<hyper_webirc>
hi vandenoever
<disasm>
graperoot: it will get garbage collected
<adisbladis>
Mic92: Needs more research then :> Just a quick glance over the original ZoL PR looks promising
<Mic92>
adisbladis: you need just a unencrypted partition, where you can store a key
<hodapp>
disasm: I'm a bit surprised at how snappy this is, and also feel a little silly that I thought the 'best' way of doing things prior was to be stuck with VirtualBox's viewer
justbeingglad has left #nixos []
<graperoot>
why isn't deluge available through nix-env -i anyway ?
<goibhniu>
graperoot: if nix-shell works for you then nix-env should also work. nix-env will install it to your user profile, so you'll have it available permanently.
<hodapp>
this is using Remmina over SPICE and thus far it seems flawless
<clever>
xd1le: what about the df -i output?
<graperoot>
goibhniu: It doesn't for me. Also, good taste in video games.
<adisbladis>
Mic92: My original thought is to have a gpg-encrypted keyfile and unencrypt that one with the smartcard
<adisbladis>
hyper_webirc: I disagree. I'm too paranoid to have keys in software ;)
<hyper_webirc>
adisbladis: or just use Correct Horse Battery Staple :)
<adisbladis>
Also work in security soo..
<hyper_webirc>
keys in software? oO
<adisbladis>
hyper_webirc: How do you know my password :O
<adisbladis>
hyper_webirc: No crypto keys on disk
<adisbladis>
Everything goes in HSMs and other secure coprocessors
<hyper_webirc>
with luks, your keys is on disk
<clever>
xd1le: ah, i think the problem is in /nix/.rw-store, but nixos-install should have chroot'd
<clever>
xd1le: have you ran nixos-rebuild?, and what args did you give to nixos-install?
<goibhniu>
graperoot: ah, does `nix-env -iA nixos.deluge` work for you?
<adisbladis>
hyper_webirc: With luks I'm using a challenge-response with a hardware key
MercurialAlchemi has quit [(Ping timeout: 248 seconds)]
<adisbladis>
So I don't actually have anything on disk. I don't even know my disk crypto keys.
<hyper_webirc>
but the actual key is still on the disk
bfrog has quit [(Quit: WeeChat 1.9.1)]
<adisbladis>
Well yes.. Sadly that cannot yet be circumvented
<hyper_webirc>
just the unlocking of the actual encryption key is not on disk ;)
<disasm>
graperoot: you tried nix-env -iA nixos.deluge (or nixpkgs if you're not using nixos)
<hyper_webirc>
I use password
tmaekawa has quit [(Quit: tmaekawa)]
<hyper_webirc>
smartcards etc are too complicated
<goibhniu>
graperoot: so, in nix we have package "names" and "attribute paths" ... the name for deluge seems to be "python2.7-deluge", so you could install that by name with `nix-env -i python2.7-deluge`
superherointj has joined #nixos
sigmundv_ has joined #nixos
<copumpkin>
installing python packages with nix-env doesn't really work afaik
<goibhniu>
graperoot: you can see the attribute paths and names with `nix-env -qaP | grep deluge`
<copumpkin>
maybe it does?
<copumpkin>
I thought they didn't end up in PYTHONPATH properly if you wanted to import them
<adisbladis>
hyper_webirc: It's super easy to replace your ssh keys with hw-backed keys :)
<copumpkin>
if they have binaries those would end up in your PATH
periklis has quit [(Ping timeout: 240 seconds)]
<hyper_webirc>
adisbladis: the problem is, you'll need to have your hw with you.... ;)
<goibhniu>
python packages should run, but you can't install them and import them as modules
<graperoot>
goibhniu: thanks a lot.
Itkovian has joined #nixos
<adisbladis>
hyper_webirc: I have multiple hw keys :) One always in the laptop, one on my keyring and then I also have ssh-keys in my phones secure element
<goibhniu>
graperoot: you're welcome! Installing by attribute path is also much faster.
<hyper_webirc>
adisbladis: I have one hw key, it's attached to my neck ;)
takle has joined #nixos
<hyper_webirc>
if I forget that one, I have other problems
* goibhniu
isn't a gamer BTW ... perhaps a gamer is impersonating me? :D
<graperoot>
that's a hell of a coincidence ! thought it was a reference to goibniu from valdis story
<xd1le>
yeah I remember doing this same installation procedure back in march and it worked..
<clever>
xd1le: you can also enable swap and resize /nix/.rw-store/, if you have the disk space for it
<xd1le>
clever: garbage collect with nix-collect-garbage -d?
<clever>
normal gc
<hyper_webirc>
or just install the bare minimum, boot into it, and build the rest?
<xd1le>
I've enabled swap with swapon if that's what you mean
<xd1le>
without -d?
<clever>
xd1le: mount /nix/.rw-store -o remount,size=5g
<xd1le>
clever: oh
<xd1le>
it is ram
<clever>
yeah, its trying to build the entire target on a ramdisk
<xd1le>
but i should have 4gb ram
<clever>
linux defaults to 50% for a tmpfs
<clever>
and it can use swap
<xd1le>
clever: do you think 5g would be enough? also this is temporary right?
<xd1le>
that makes sense
<clever>
yeah
<hyper_webirc>
btw, how would I directly use nixos-unstable channel when installing a new system? last time I tried in the live usb to switch channels but it still didn't use the unstable one
superherointj has quit [(Quit: Leaving)]
<graperoot>
nixos is honestly really awesome
<xd1le>
hyper_webirc: (that's why I'm using -I to do it, but idk if there's a better way)
<hyper_webirc>
I just thought if I switch to unstable channel on the live usb it would work
<xd1le>
yeah I would think so too :/
<hyper_webirc>
well, it didn't for me... but that could also just be pebkac
<xd1le>
because with -I you'd at least have to download an entire archive of nixpkgs or something
katyucha has quit [(Ping timeout: 248 seconds)]
<hyper_webirc>
I'll just do a minimum install, boot into it, switch channels and build all the rest :)
<xd1le>
clever: thank you so much. nixos-install is currently back to downloading stuff so I assume it's going as planned again. :)
Ralith_ has joined #nixos
erasmas has joined #nixos
<xd1le>
hyper_webirc: yeah can always fallback to that
<xd1le>
but not the exact answer to your question
leat has quit [(Quit: WeeChat 1.9)]
<xd1le>
but damn this limitation really sucks :/
<xd1le>
oh wells
<hyper_webirc>
if life was easy .... :)
<xd1le>
hyper_webirc: yeah basically i'm too lazy to do a minimal install the first time and then do the bigger one after booting into to it. just wanna do everything at once :P
<hyper_webirc>
(you could always hire someone to do it for you...)
<xd1le>
hyper_webirc: because a minimal install means probably editing the configuration.nix a bit, instead of just pulling it from git
<xd1le>
or
<hyper_webirc>
nothing really to edit in the minimal install configuration.nix
<xd1le>
i could have a minimal configuration.nix in git
<xd1le>
oh nice
<hyper_webirc>
(except for me adding zfs)
<hyper_webirc>
and adding according keyboard layout
<clever>
graperoot: nix builds never see anything you install with nix-env
<xd1le>
hyper_webirc: clever: wow this is so nice!
<clever>
graperoot: they are all sandboxed
<graperoot>
Wooot
<graperoot>
how do I solve this ?
<clever>
graperoot: you need to create an override against the package that is failing to build
katyucha has joined #nixos
<clever>
graperoot: to start with, what package did nix say failed to build?
<xd1le>
clever: leksah
<graperoot>
build of ‘/nix/store/bsq7jix2kwp8rk72wf36mzhgld5818cr-leksah.drv’ failed
pie_ has quit [(Ping timeout: 248 seconds)]
<hyper_webirc>
xd1le: leksah?
<clever>
graperoot: was there anything else on that line?
<clever>
graperoot: it normally has an exit code after the word failed
<graperoot>
Oh ! sorry. yes.
vandenoever has quit [(Ping timeout: 258 seconds)]
<graperoot>
Encountered missing dependencies:
<graperoot>
haskell-gi-overloading ==0.0.*
<graperoot>
builder for ‘/nix/store/9ik1wfa27sjx4ca8l2q650wjamnff784-ltk-0.16.1.0.drv’ failed with exit code 1
<graperoot>
cannot build derivation ‘/nix/store/01hrmnqh3630jbq9az5hzm02qxgmdk3h-ghc-8.0.2-with-packages.drv’: 1 dependencies couldn't be built
<graperoot>
cannot build derivation ‘/nix/store/bsq7jix2kwp8rk72wf36mzhgld5818cr-leksah.drv’: 1 dependencies couldn't be built
<xd1le>
clever: I get a "error: path '/nix/store/<hash>-nixos-system-nixos-18.03pre117013.aebdc892d6' does not exist and cannot be created" error for that nixos-install.. any ideas?
leinadlime has quit [(Quit: Connection closed for inactivity)]
<clever>
graperoot: so leksah never even started to build, the problem is over in ltk
<clever>
xd1le: your -I is pointing to a path that only exists on the host, use nix-channel instead
<clever>
xd1le: double-check with nix-channel --list that you gave it the right name, and there is only one channel
Neo-- has joined #nixos
<graperoot>
yep..
<clever>
i can confirm the same problem on this end
<xd1le>
okay so what i did was -I nixos-config=/mnt/home/u/dots/nixos/config.nix -I nixpkgs=/mnt/home/u/dots/nixos/nixpkgs ..
corpix has quit [(Remote host closed the connection)]
<nixer>
I have two modules that import one module, I somtimes enable one of the modules or both however when I enable both modules I get an error that systemd.services.xxx.serviceConfig.Type is defined multiple times. Is there some way I can magically de-dupe imports or merge the config?
corpix has joined #nixos
oida_ has joined #nixos
goibhniu has quit [(Ping timeout: 260 seconds)]
oida has quit [(Ping timeout: 260 seconds)]
mudri has joined #nixos
phreedom has quit [(Ping timeout: 246 seconds)]
taaperotassu has quit [(Ping timeout: 248 seconds)]
<csingley>
nixos-rebuild creates hmc93i8ph2l2x783cj4bs8q1rrdcklkg-etc-vimrc with correct contents
<csingley>
However /etc/vimrc doesn't get linked, and therefore not loaded when vim starts.
<csingley>
What am I doing wrong?
<cement>
are you copying what's in that guide or did you change things?
JagaJaga has quit [(Ping timeout: 255 seconds)]
<csingley>
No matter what I put for vim_configurable.customize.vimrcConfig.customRC, it goes into the nix store, and never gets linked to /etc/vimrc
<cement>
hmmm
<LnL>
vim_configurable doesn't create anything in /etc
<csingley>
Oh. So how's that directive supposed to work?
<csingley>
There is an etc-vimrc being created; I just assumed...
<LnL>
the vim resulting binary will load the /nix/store/hmc93i8ph2l2x783cj4bs8q1rrdcklkg-etc-vimrc
<copumpkin>
nix builds can never touch global locations like /etc
<nixer>
ryantm: Thansk.
<copumpkin>
we use other trickery to convince tools to look in non-global places
<hyper_webirc>
ah, didn't make it bootable :)
MichaelRaskin has joined #nixos
<csingley>
Thank you. Obv a new user. So: why isn't the vim binary loading my configs (which are created correctly)?
<LnL>
are you sure it's not getting loaded?
<cement>
something I had to do was nix-env -e vim
<LnL>
it should run whatever you add to customRC however it won't look at ~/.vimrc by default IIRC
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] woffs opened pull request #30090: devtodo: init at 0.1.20 (master...devtodo1) https://git.io/vd4OO
NixOS_GitHub has left #nixos []
<csingley>
Configs aren't getting loaded. For a quick test I turn on line numbering; obvious if not honored.
<csingley>
Good idea uninstalling plain vim (I had it installed before), but alas it's not even in /nix/store (garbage collected)
<csingley>
Anybody know how to check from within vim what configs it's loaded?
<LnL>
how did you install it?
<cement>
did you change the name = line in the example from that guide?
<csingley>
I installed vim_configurable to the system configuration.nix. So far no software installed to the user profile
cpennington has joined #nixos
<csingley>
I set name = "vim"
<cement>
hm
<LnL>
so with environment.systemPackages? what does type -P vim show?
<csingley>
1) Yes sir. 2) "vim is /run/current-system/sw/bin/vim"
<LnL>
as for seeing if it's loaded, adding a syntax error is probably the easiest way to check :)
MP2E has joined #nixos
<LnL>
all of that looks correct so I would expect it to just work
boxofrox has quit [(Ping timeout: 248 seconds)]
<hyper_webirc>
so, doing now the big install...
<csingley>
`vim -u hmc93i8ph2l2x783cj4bs8q1rrdcklkg-etc-vimrc $file` works fine
<cement>
did you directly enter your vimrc into the nix file, or are you loading it from a file?
<csingley>
cement: I have tried entering the config directly into configuration.nix, and loading it from a separate module. neither works
_ris has joined #nixos
<csingley>
Since I can manually specify the correct config file (as built by nixos-rebuild), and it passes, I'm thinking the problem is not in my vim syntax or my nix config expression
<csingley>
But everybody else seems to be using it just fine?
<csingley>
This is on an updated 17.09 channel BTW
<LnL>
what about file $(which vim)
<csingley>
"command not found" LOL. zsh
<LnL>
nix-shell -p file which --run 'file $(which vim)'
infinisil has joined #nixos
a6a3uh has joined #nixos
<csingley>
There we go: "/run/current-system/sw/bin/vim: symbolic link to /nix/store/qi7n9xamd2gw2hzq5z3izj7mxskqylyy-vim-8.0.1150/bin/vim"
reinzelmann has joined #nixos
<LnL>
well yes :) but the actual file, is that a script or a binary?
<csingley>
binary "/nix/store/qi7n9xamd2gw2hzq5z3izj7mxskqylyy-vim-8.0.1150/bin/vim: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /nix/store/xzx1bv1d7z4mgg6sg6ly0jx609qvka4x-glibc-2.25-49/lib/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, stripped"
<LnL>
ok, that's not your vim_configurable then
taaperotassu has joined #nixos
boxofrox has joined #nixos
<cement>
how did you install vim again?
<csingley>
Must not be. I do wish it were, though. My config file appears to declare it so.
<LnL>
I'm suspecting you have multiple vim packages in systemPackages
some-user has joined #nixos
<some-user>
hello
<some-user>
i've installed nixos 17.09 and gnupg, transferred .gnupg from my other box and nixos box says it can't find public key during file.gpg decrypting
zzamboni has joined #nixos
<some-user>
what's wrong?
zzamboni has quit [(Client Quit)]
<csingley>
I think the problem may be that I set programs.vim.defaultEditor = true
<csingley>
That may be pulling in the wrong vim
zzamboni has joined #nixos
<csingley>
I'm going to try rebuilding without that.
bebarker has quit [(Remote host closed the connection)]
nil has quit [(Ping timeout: 258 seconds)]
<sophiag>
I'm using Nix on Darwin for the first time today (normally use NixOS with no separate config.nix) and am getting an odd error trying to install an environment for the first time. It seems linux/systemd related...which is throwing me for a loop. Here's the short trace: https://gist.github.com/Sophia-Gold/3bf3594cf476650414977c862a4d8598
<csingley>
My vim is now correctly configured, and the world is a better place.
<csingley>
LnL, cement: thank you for helping me
<hyper_webirc>
so, duing full install now and restoring data from backup :)
zzamboni has quit [(Quit: Leaving.)]
boxofrox has quit [(Ping timeout: 260 seconds)]
<cement>
you're quite welcome. I literally just spent 2 work days and some 6-8 hours setting up my own vim through nix
<LnL>
you have allowBroken = true; systemd is a package that doesn't work on darwin
nil has joined #nixos
<sophiag>
Lnl: ah, I wouldn't have thought broken would be used for that. Let me try without it
<csingley>
This host has another small glitch I wonder if anyone can illuminate
<csingley>
I have a nixos container with interfaces = [ "eno2" ] declared, to pass through a dedicated enet i/f
<LnL>
allowBroken makes nix ignore meta.platforms so stuff that's can fail in confusing ways like that
<csingley>
This seems to break systemd-dhcpcd, so I have to statically configure IP
orivej has quit [(Remote host closed the connection)]
orivej has joined #nixos
<csingley>
Not a big deal, but I have done this same configuration for systemd-nspawn containers in other Linux distributions with no problem
<csingley>
I wonder if the networking setup doesn't account for this case; has anyone got DHCP working on a network i/f passed through to a systemd-nspawn container in NixOS?
<hyper_webirc>
Mic92: still here?
<sophiag>
LnL: The reason I set allowBroken to true was because it's throwing an error for Chromium when I don't even have it included as a package (I use Chrome unfree). Is it a dep for that?
<LnL>
don't think we have a chrome for darwin
<LnL>
not even sure if there's any browser that works
<gchristensen>
w00t my nixos hummingbird shirt arrived!
<sophiag>
Turns out Chromium is a dep for Chrome. I'm fine if I need to install stuff like that by hand as long as my dev tools work
<infinisil>
some-user: no idea, you could try reimporting the keys maybe
<jsgrant_>
gchristensen, Where do you get NixOS "swag"?
<jsgrant_>
Ughh ... remembered I still need to figure out the branch-issue today or tomorrow; 17.09 on my unstable box & 18.03 on all my 17.09 boxes. :^U
<__monty__>
Why was the hexagon rotated?
<jsgrant_>
Reservation made; Maybe someday in a few-months-to-year I'll be happily surprised with a 20 dollar teespring payment I forgot I made/convienced myself that my card was cracked.
Mateon1 has quit [(Ping timeout: 248 seconds)]
<gchristensen>
__monty__: perhaps a better question is "why was a new logo designed?"
schoppenhauer has quit [(Ping timeout: 258 seconds)]
<__monty__>
I figured I'd be specific about what I wanted to know ; )
<gchristensen>
__monty__: the hexagon was not rotated, a new logo was designed and chosen
<__monty__>
But, the hexagon was rotated.
* jsgrant_
is now back to NixOS on all my non-phone/tablet/rpi boxen; Planning to stay.
zzamboni has quit [(Quit: Leaving.)]
schoppenhauer has joined #nixos
<jsgrant_>
Going to be weird to learn Ada in/with it though, have done such trivial dev thusfar in NixOS it's embrassing.
<jsgrant_>
A whole lot of config.nix fiddling though.
<__monty__>
gchristensen: Was it? Looks like the spacing was just changed.
<jsgrant_>
__monty__: Both?
Mateon1 has joined #nixos
<jsgrant_>
Almost (maybe all) letters are wider.
qrvaelet has joined #nixos
<__monty__>
I see it for the E's and the T but I'm still not sure why it's relevant. If the G is wider it's obviously very subtle. The rotation of the hexagon makes a much bigger difference.
pie_ has quit [(Remote host closed the connection)]
pie_ has joined #nixos
* jsgrant_
didn't notice the rotation till mentioned, but did notice spacing was wider -- subjective, but also guess there's no side-by-side.
<jsgrant_>
Image for nixos*
moet has joined #nixos
mingc has quit [(Remote host closed the connection)]
reinzelmann has quit [(Quit: Leaving)]
muffins_Madcat has joined #nixos
<__monty__>
Think about encountering them seperately. You see so many fonts in one day, would you really notice some different kerning? On the other hand the nixos logo went from jagged and more snowflake-like to straight edges.
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<qrvaelet>
I'd like to get some help for my nixos, since after upgrade the system is broken. My current generation uses 4.13.2 kernel but on 4.13.4 or 4.14-rc2 I cant use any graphical application, when I try to do it Xorg/everything hangs but I can move my mouse cursor. I'm on unstable. Where can I look after this issue?
zzamboni has joined #nixos
GrassAs has joined #nixos
<GrassAs>
FYI, wineStaging pulls samba3 which is tagged insecure.
GrassAs has quit [(Client Quit)]
<moet>
hey nixos, what's your favorite email reader?
<jsgrant_>
__monty__, Probably actually -- but I'm more focused on spacing than many it seems.
<jsgrant_>
So, don't think I'm 'the norm'.
<jsgrant_>
moet, A lot of people -- generally like Nylas.
mortum5 has joined #nixos
mortum5 has quit [(Max SendQ exceeded)]
mortum5 has joined #nixos
mortum5 has quit [(Max SendQ exceeded)]
<jsgrant_>
In-general, don't know about NixOS users; I personally think 'Thunderbird' is still good-enough for now/for the most part.
mortum5 has joined #nixos
<moet>
ohh, i hadn't seen nylas
<moet>
looks nice
<moet>
i've used thunderbird in the past
<qrvaelet>
nylas is some electron based application
<__monty__>
Yeh, terrible.
<moet>
i'm looking for something maybe more minimal, but just thought i'd ask in this channel
<moet>
my computer is light on ram, heh
<qrvaelet>
claws maybe
<__monty__>
Only thing nylas does well is visual design. Thunderbird is superior in all else.
<qrvaelet>
or geary
<gchristensen>
I like notmuch.el :)
tester123 has joined #nixos
zzamboni has quit [(Quit: Leaving.)]
<__monty__>
No problem with thunderbird with 2G ram.
zzamboni has joined #nixos
<tester123>
Has anyone used SteamVR with NixOS?
<tester123>
I'm getting I'm getting error "Make sure your headset is connected properly, then reboot it by clicking below". I installed `steam` from nixpkgs. I'm worried the issue is connected with the fact that you have to have certain udev rules for SteamVR to work: https://github.com/ValveSoftware/SteamVR-for-Linux
<qrvaelet>
__monty__: i worked in support, I had thousands of mails in my inbox, thunderbird easily ate up 1-2Gigs of memory, but luckily thats not a general use case
<__monty__>
qrvaelet: Did claws/geary handle it better?
<tester123>
That said I'm tempted to restart again to see if the problem goes away.
glaebhoerl has joined #nixos
<csingley>
You can try systemctl restart systemd-udevd first
qrvaelet has joined #nixos
<qrvaelet>
meh
<qrvaelet>
no luck :(
<qrvaelet>
I'm on the non-working generation
<csingley>
If that doesn't work, I'd try `lsusb -v` to pull idVendor & idProduct for the Vive components
<csingley>
See how they match with the rules you're applying
<qrvaelet>
where should I look to identify the responsible module/package for my Xorg hangs?
a6a3uh has quit [(Quit: Bye)]
peacememories has joined #nixos
<glaebhoerl>
hello. I added 3 of the new services listed in the 17.09 release notes (powertop, qt5ct, thefuck) to my configuration.nix, added the 17.09 channel and updated it, and did `nixos-rebuild boot`, and for all 3 of them (if I comment out the previous ones of course), it complains e.g. "error: The option `powerManagement.powertop' defined in `/etc/nixos/configuration.nix' does not exist."
<glaebhoerl>
the fact that it's for all 3 of the new ones makes me suspect it's not just some coincidence and I might be doing something stupid, but right now I can't see what
digitus has joined #nixos
<glaebhoerl>
I basically just did `nix-channel --add https://nixos.org/channels/nixos-17.09 nixos; nix-channel --update nixos; nixos-rebuild boot;`, and the options are specified as `powerManagement.powertop.enable = true; programs.qt5ct.enable = true; programs.thefuck.enable = true;`
<glaebhoerl>
does anyone see anything wrong here, or has run into anything similar?
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] zzamboni opened pull request #30094: vagrant: fix Ruby and libffi libraries on Darwin (master...vagrant-fix-darwin) https://git.io/vd420
NixOS_GitHub has left #nixos []
maiksen has joined #nixos
<mingc>
glaebhoerl: do you run nix-channel command as root?
<glaebhoerl>
yes, all of the commands were as root
<glaebhoerl>
conversely, can anyone confirm that any of these 3 options _do_ work for them on 17.09?
tester123 has quit [(Ping timeout: 260 seconds)]
tester123 has joined #nixos
<tester123>
SteamVR works now! Except I'm getting these new errors: "Error: You are missing the following 32-bit libraries, and Steam may not run: libgobject-2.0.so.0, libglib-2.0.so.0, libgio-2.0.so.0, libgtk-x11-2.0.so.0, libpulse.so.0, libgdk_pixbuf-2.0.so.0"
zzamboni has quit [(Quit: Leaving.)]
<csingley>
tester123: nice. Do you have "hardware.opengl.driSupport32Bit = true; hardware.pulseaudio.support32Bit = true;" ?
<mingc>
glaebhoerl: I just tested powertop.enable, and it works for me on 17.09
<glaebhoerl>
mingc: thanks!
jb55 has joined #nixos
taaperotassu has joined #nixos
<tester123>
csingley: Yes.
oida_ has joined #nixos
<glaebhoerl>
I get the feeling that for some reason it still wants to use the old channel or something, is there anything I could've done previously to cause that? and/or any way I can check if that's the case? (I did see it downloading the 17.09 channel, multiple times, and it's there in the list, and the hash seems good, so presumably that's not the issue...)
jtojnar has quit [(Remote host closed the connection)]
<glaebhoerl>
ah ok, I fixed it :P not surprisingly it was because of a cute local hack I have to have a user called "su" aliased to uid 0. apparently nix was storing the channels variously under /nix/var/nix/profiles/per-user/su/ or /nix/var/nix/profiles/per-user/root/ and confusing itself.
cpennington has quit [(Remote host closed the connection)]
<csingley>
tester123: too bad. The nix pkg for steam utilizes the steam runtime I believe; can you confirm that from the script output when you run steam?
peacememories has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
cement has quit [(Ping timeout: 258 seconds)]
<tester123>
csingley: What do you mean?
<pingveno>
I'm having some trouble with "error: unexpected EOF reading a line" when attempting to install or upgrade some new packages. From a nix-env --upgrade run: https://clbin.com/LtCjl
<pingveno>
I was able to upgrade every other derivation in the user environment except the ones listed and GCC
<tester123>
I'm getting ` SteamVR failed to initialized for unknown reasons. (Error: Shared IPC Compositor Connect Failed (306))`
<tester123>
and `sh: lsb_release: command not found` and `sh: pidof: command not found` as well as the ` Error: You are missing the following 32-bit libraries, and Steam may not run: libgobject-2.0.so.0, libglib-2.0.so.0, libgio-2.0.so.0, libgtk-x11-2.0.so.0, libpulse.so.0, libgdk_pixbuf-2.0.so.0`
<Eisfreak7>
I just switched to 17.09 and it broke my terminal fonts. I read that there was a switch in font configuration, but that shouldn't be breaking right? It appears like my terminal emulator (termite) can't find my font anymore and just uses the default one
Itkovian has joined #nixos
JosW has quit [(Quit: Konversation terminated!)]
sigmundv__ has joined #nixos
hiratara has quit [(Ping timeout: 246 seconds)]
civodul has quit [(Quit: ERC (IRC client for Emacs 25.3.1))]
hiratara has joined #nixos
<kkini>
I guess Nix derivations don't always turn out to be 100% deterministic, do they? So on two different machines, building the same nix store path might result in (slightly) different data, for example if the builder script reads /dev/urandom or whatever.
<kkini>
If that's the case, then how can a user independently verify the stuff that they download from a remote cache?
<MoreTea>
Yes in theory. Other sources of indeterminism can be e.g. timestamps included by compilers
<MoreTea>
You can run nix-build with the --check option to rebuild and see if the output is exactly the same as one that has been build/downloaded before.
<MoreTea>
Also see the build-repeat configuration option, which repeats building a derivation a number of times, and only puts it in the store if all of the rebuilds are identical
<kkini>
I see. (And I guess that `nix-build --check` will often fail, then.)
<kkini>
Then I guess the signatures on cached objects aren't meant to be attesting to the fact that the cached object would pass `nix-build --check`. They're just attesting to the fact that they were produced by some trusted agent (e.g. hydra.nixos.org).
<MichaelRaskin>
Yes
<MoreTea>
yes
<kkini>
I see, I guess I was misunderstanding that :) Thanks for the clarification!
<MoreTea>
however I've read some positive sounds that actually quite some packages _are_ reproducible
<MoreTea>
kkini, you pretty much have absolute control over which software is brought into scope for building some other program, so by default nix is doing pretty good already.
Itkovian has joined #nixos
<kkini>
Yup. It's definitely miles ahead of any other distro / package manager I've seen, in terms of reproducibility :)
<sophiag>
i'm having some trouble with emacs (with packages) installed on os x. i have the legacy os x version as well as 25 from nix and the gui version (presumably the old one) cannot locate the packages nix installed
python476 has joined #nixos
dejanr has quit [(Ping timeout: 248 seconds)]
sigmundv has quit [(Client Quit)]
zzamboni has quit [(Quit: Leaving.)]
iqubic has quit [(Ping timeout: 260 seconds)]
ertes has joined #nixos
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
Acou_Bass is now known as eddie
eddie is now known as Acou_Bass
maiksen has quit [(Ping timeout: 260 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] jtojnar opened pull request #30102: simple-scan: Move to GNOME package set (master...gnome-simple-scan) https://git.io/vd45N
NixOS_GitHub has left #nixos []
digitus has quit [(Quit: digitus)]
dejanr has joined #nixos
moet has joined #nixos
dywedir has quit [(Remote host closed the connection)]
slyfox has quit [(Ping timeout: 240 seconds)]
slyfox has joined #nixos
erasmas has quit [(Quit: leaving)]
griff_ has quit [(Quit: griff_)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] primeos pushed 1 new commit to master: https://git.io/vd4bs
<NixOS_GitHub>
nixpkgs/master c42b878 Michael Weiss: fscrypt: 0.2.1 -> 0.2.2
NixOS_GitHub has left #nixos []
<sophiag>
so it appears Nix isn't creating symlinks for my emacs packages correctly under Darwin. can anyone help with this?
<sophiag>
i only have git.el and git-blame.el in ~/.nix-profile/share/emacs
<sophiag>
the only way emacs locates the rest is if i launch it inside a terminal
moet has quit [(Quit: leaving)]
dieggsy has joined #nixos
<sophiag>
anyone? #emacs folks are being very rude about this unfortunately :/
mudri has quit [(Ping timeout: 248 seconds)]
bennofs has quit [(Ping timeout: 246 seconds)]
<alp>
sophiag, are you aware of the 'emacsWithPackages' function?
<alp>
that would get emacs and the emacs packages through nix, but it's probably the easiest way to get to something that works
Acou_Bass is now known as eddie
eddie is now known as Acou_Bass
<alp>
alternatively, you have to hunt down the paths of all your emacs packages in the nix store and add those in your emacs config
acarrico has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] dtzWill opened pull request #30103: lsb-release: wrap to ensure needed utilities are available (master...fix/lsb-release-PATH) https://git.io/vd4AU
NixOS_GitHub has left #nixos []
markus1199 has joined #nixos
<sophiag>
alp: i'm using emacsWithPackages
jb55 has quit [(Ping timeout: 248 seconds)]
<sophiag>
i've been doing that on NixOS at home and never had trouble. now having this problem on Darwin at work
<sophiag>
to be clear my config.nix has: "(emacsWithPackages (with emacs25PackagesNg; [ .. ]))"
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
markus1189 has quit [(Ping timeout: 260 seconds)]
<alp>
sophiag, are you sure the emacs you're running is the one from nixpkgs?