silver has quit [Read error: Connection reset by peer]
Edes has joined #nixos
rajivr___ has joined #nixos
orivej has quit [Ping timeout: 252 seconds]
<Edes>
Does anyone know how to get emoji working properly on nixos? I've been poking around with the fontconfig options, but I haven't been able to get it to render all emoji properly, the problem is that the emoji from another font (I think it's roboto) are getting rendered in place of noto.
<infinisil>
> options "services.xserver.libinput.horizontalScrolling" # <-- Ahh, there we go, for future reference
<{^_^}>
"Disables horizontal scrolling. When disabled, this driver will discard any horizontal scroll\nevents from libinput. Note that this does not disable horizontal scrolling, it merely\ndiscards the horiz...
<hexa->
haha, nice
Guanin has quit [Remote host closed the connection]
myskran has quit [Ping timeout: 244 seconds]
is_null has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
Synthetica has quit [Quit: Connection closed for inactivity]
camsbury has joined #nixos
ddellacosta has quit [Ping timeout: 246 seconds]
o1lo01ol1o has joined #nixos
mexisme has joined #nixos
i1nfusion has quit [Remote host closed the connection]
i1nfusion has joined #nixos
myskran has joined #nixos
myskran has quit [Ping timeout: 244 seconds]
lunik1 has quit [Ping timeout: 246 seconds]
lunik1 has joined #nixos
thomashoneyman has quit [Ping timeout: 252 seconds]
i1nfusion has quit [Remote host closed the connection]
i1nfusion has joined #nixos
i1nfusion has quit [Remote host closed the connection]
i1nfusion has joined #nixos
toppler`` has quit [Ping timeout: 268 seconds]
hamishmack has joined #nixos
toppler`` has joined #nixos
i1nfusion has quit [Remote host closed the connection]
i1nfusion has joined #nixos
iqubic has quit [Quit: Page closed]
mexisme has quit [Ping timeout: 246 seconds]
xkapastel has joined #nixos
lightbulbjim has quit [Max SendQ exceeded]
lightbulbjim has joined #nixos
hamishmack has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
dycan has joined #nixos
esmerelda has quit [Ping timeout: 268 seconds]
lambda-11235 has joined #nixos
esmerelda has joined #nixos
o1lo01ol1o has quit [Ping timeout: 246 seconds]
jackdk_ has joined #nixos
hamishmack has joined #nixos
o1lo01ol_ has joined #nixos
jackdk_ has quit [Ping timeout: 246 seconds]
myskran has joined #nixos
myskran has quit [Ping timeout: 258 seconds]
jackdk_ has joined #nixos
abathur has joined #nixos
camsbury has quit [Remote host closed the connection]
noam_ has joined #nixos
noam has quit [Ping timeout: 258 seconds]
<colemickens>
anybody have good ideas how to handle plugins for OBS in nixpkgs?
o1lo01ol_ has quit [Remote host closed the connection]
abathur has quit [Ping timeout: 246 seconds]
abathur has joined #nixos
Supersonic has quit [Ping timeout: 252 seconds]
abathur has quit [Ping timeout: 244 seconds]
drakonis has quit [Quit: WeeChat 2.4]
InFlames has joined #nixos
Supersonic has joined #nixos
<{^_^}>
[nixpkgs] @lilyball opened pull request #61483 → unison: fix building on macOS with enableX11 = false → https://git.io/fjWdJ
<sphalerite>
Having an interesting problem… `mount /dev/mmcblk1p1 /boot` doesn't complain, and the kernel log says EXT4-fs (mmcblk1p1): mounted filesystem with ordered data mode. Opts: (null)
<sphalerite>
but it still doesn't seem to actually mount it!?
<sphalerite>
i.e. it doesn't appear in /proc/mounts, nor does lost+found appear in /boot
knupfer has joined #nixos
<sphalerite>
any ideas?
woffs has quit [Quit: Gateway shutdown]
<{^_^}>
[nixpkgs] @FRidh pushed 5 commits to staging-19.03: https://git.io/fjWbN
<{^_^}>
[nixpkgs] @FRidh pushed 88 commits to staging-19.03: https://git.io/fjWbA
woffs has joined #nixos
mexisme has quit [Ping timeout: 258 seconds]
<clever>
sphalerite: what does `strace mount` report?
hoijui has joined #nixos
<sphalerite>
clever: strace mount, or tracing the mount command that mounts the filesystem?
<sphalerite>
I was going to say, maybe I should use that for my init script as well instead of sourcing the various uevent files in /sys/class/block and using mknod… but then I realised I don't use that script anymore, thanks to modern u-boot \o/
<clever>
sphalerite: i once wrote bash scripts to deal with firmware loading
<clever>
so i could get wifi in my custom initrd
<sphalerite>
pfff who needs bash when you have busybox ash :p
<clever>
basically, you need to write the size of the firmware to a special /sys entry, then the blob itself, and do it in response to a hotplug event thing
noam has joined #nixos
<clever>
and either ack or nak depending on if you find it
<rauno>
gchristensen, one more question about theat ixpe and packet setup, do you know how their metadata service works? like which data should be in response ?
<ivan>
ah I finally figured out why NVIDIA/Xorg didn't work for me in master; the new graphene-hardened allocator enabled by default in the hardened profile
<Taneb>
clever: OK. Plenty of spare CPU and mem, not using swap
<clever>
Taneb: find the pid's for nix-serve, and run `strace -p pid1 -p pid2 ...` and then see what happens when you curl it
<yorick>
clever: 192.80.8.124
<yorick>
there's no route for it though, it would go via the public router
<clever>
ah, i was thinking about how aws's hypervisor handles things, but packet is baremetal
cfricke has joined #nixos
jomik has joined #nixos
<Taneb>
clever: ...it seems to have fixed itself
<jomik>
Why can I not give a `nix-env` call the argument `-I '<unstable>'`? It says unstable does not exist, but `nix-channel --list` does show it..
<clever>
jomik: -I cant take search paths like <foo>
fendor has joined #nixos
<jomik>
Ahhhh. I just got an issue that said it used that :D
<jomik>
Thanks.
<__monty__>
jomik: In this case just use -f rather than -I.
<jomik>
__monty__: That gives `error: access to path '/nix/store/b23jd9xdjj3sav7jv2bss448vzj1xjls-unstable-19.09pre178675.24debf74ef5/unstable' is forbidden in restricted mode`
hoijui has joined #nixos
<clever>
jomik: why is restricted mode enabled?
<jomik>
Trying to replicate the NUR call. My NUR expressions are failing in their CI :D
<clever>
jomik: and you shouldnt be using nix-env when doing testing or ci type things
<clever>
that will ruin your disk usage
<clever>
just use plain nix-build
<jomik>
Hmmm. Lol, true. That would install all the things XD
<jomik>
Thank you, saved me :D
<jomik>
How would I test that CI call that they are doing for NUR?
<clever>
--option restrict-eval true
<clever>
jomik: that is why restricted eval was on
lovesegfault has quit [Ping timeout: 252 seconds]
<jomik>
Yah. I don't know why they have that there though.
<clever>
and they are doing -qa, which nix-build doesnt support
Guest91665 has left #nixos ["User left"]
lovesegfault has joined #nixos
MmeQuignon has joined #nixos
hoijui has quit [Ping timeout: 258 seconds]
ixxie has joined #nixos
<ixxie>
hello folks
tsmeets has joined #nixos
avn has joined #nixos
<jomik>
I am trying to resolve this issue __monty__ clever https://gitlab.com/Jomik/nur-expressions/issues/1 If you have any idea how, then that would be awesome. I am trying his fix, but that does not seem to be able to build :/
<Taneb>
Aaargh, the thing I thought fixed itself has unfixed itself, my nix-serve is once again slow
<yorick>
jomik: we fixed your issue in npm-buildpackage, btw ;)
MmeQuignon has quit [Ping timeout: 258 seconds]
drakonis has joined #nixos
<jomik>
yorick: I saw. I commented and updated already :P
<jomik>
Thanks alot :D
<tsmeets>
Hello, I am trying to compile a static version of libX11 (nix build nixpkgs.pkgsStatic.xorg.libX11), but it fails at libxcb with "relocation R_X86_64_32 against hidden symbol `__TMC_END__' can not be used when making a shared object"
<tsmeets>
any ideas?
<jomik>
yorick: But apparently my use of your package to build my packages is messing up the CI for NUR :D
<Taneb>
...trace is having a lot of "select(8, [5], NULL, NULL, {tv_sec=1, tv_usec=0}) = 0 (Timeout)"
acarrico has joined #nixos
<clever>
Taneb: thats normal, its waiting up to 1 second for events on fd 5
<Taneb>
clever: I'm seeing almost nothing but that
<clever>
Taneb: you can -e 'trace=!select' to hide that
<yorick>
jomik: hydra might be running it in restricted eval mode
<clever>
yorick: hydra is always in restricted mode
<jomik>
Yah, that is what they do. Do wonder how I fix it.. :(
<yorick>
clever: is fetchTarball with a hash allowed in restricted eval mode?
<yorick>
also, is import-from-derivation allowed in restricted eval mode?
<clever>
yorick: ifd is allowed, but the inputs are still bound to the normal sandbox rules
<clever>
yorick: fetchTarball requires it to be in the allowed-uris option in nix.conf
<clever>
yorick: <nixos-unstable> isnt quoted, so thats shell redirection
<yorick>
well, that would result in weirdness
hoijui has joined #nixos
drakonis has quit [Ping timeout: 246 seconds]
gal_bolle has joined #nixos
<yorick>
jomik: are you testing this with that broken nix-env?
MmeQuignon has quit [Ping timeout: 246 seconds]
<jomik>
yorick: No. Because that doesn't work xD I don't actually know how to test it... But I wrote the nix expression in that gist I linked. And that does not even build with `nix-build ci.nix -A buildOutputs` which is what my CI uses.
hoijui has quit [Ping timeout: 258 seconds]
sb0 has joined #nixos
<sb0>
I have a nix store path that has been recently created on a binary substituter, but "nix-store -r" fails to find it (as its cache contains that it does not exist on the substituters)
<cizra>
Any xmonad users around? I can't figure out how to rebuild it (is it rebuilt automatically during nixos-rebuild switch? am I supposed to rebuild myself? If so, how do I set up the right env that'd include the extra and contrib packages?
<jomik>
Well, that's the file with the dependencies I needed for my IDE to check my xmonad.hs.
orivej has joined #nixos
<jomik>
For rebuilding, I just ran the recompile command `xmonad --recompile` IIRC
<jomik>
But yes, running `nix-shell` with that shell.nix gives you the dependencies you need.
<cizra>
jomik: Yeah, it rebuilt fine inside that shell.
noam_ has joined #nixos
anderslundstedt has quit [Quit: leaving]
<cizra>
jomik: Thanks a lot. Maybe it should be documented somewhere.
Cale has quit [Ping timeout: 252 seconds]
anderslundstedt has joined #nixos
srid has joined #nixos
noam__ has quit [Ping timeout: 268 seconds]
gxt has quit [Ping timeout: 258 seconds]
rauno has quit [Ping timeout: 268 seconds]
<Lears>
cizra: You don't need to do anything special; if you just use the configuration.nix options exposed for xmonad it will be given a ghc with the right libs to recompile your config.
<jomik>
Yeah. The shell.nix I linked is just needed for your IDE to be able to find the types and such :D
<lejonet>
So what is the correct way to enable ia32 emulation for the kernel? driSupport32Bit requires it, and I have enabled it through boot.kernelPatches and adding it through extraConfig, but that doesn't flip the right attributes :P
<clever>
lejonet: the 64bit kernel has 32bit support enabled by default
<clever>
lejonet: have you changed your boot.kernelPackages?
<lejonet>
clever: yeah, I'm using the hardened kernel package
<lejonet>
Hence my asking, how do I get nix to accept that I have IA32 emulation?
<lejonet>
I've tried libMkforcing the attribute that I think driSupport32Bit without any success, and that is most likely not the right way anyway
<lejonet>
Maybe there is some overridable thing I can give it?
<Lears>
cizra: xmonad will recompile itself in the same way it does on any other distribution. You can trigger that with mod+shift+q (iirc) or `xmonad --recompile && xmonad --restart`.
<tobiasBora>
Hello,
acarrico has quit [Ping timeout: 245 seconds]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<cizra>
Lears: I think that's not the case as I'm not installing GHC explicitly.
<cizra>
Regardless, jomik's script works well.
<Lears>
Like I said, nixos will give it the ghc it needs.
<tobiasBora>
I have a quite regular "issue" with nix, which is the time to apply a simple command. For example, when I want to create a new user, it takes 5 seconds to modify the file... But "nixos-rebuild test" (or switch) takes ages, just to run the first 2 steps "building Nix...building the system configuration..." (I lose maybe 2mn before seing the first "these derivations will be built"). I'm running a recent NixOs,
<tobiasBora>
and I'm wondering if there is no way to improve this time. For what I understand, this time is long because it takes time to load all the modules. But couldn't it be possible to run a daemon that load these modules once? I'm tired of waiting 5mn (literraly, since my hello the command has not stopped yet) at every change, and NixOs is so nice that I don't want to be upset against him for such a stupid reason ^^
<Baughn>
tobiasBora: Possible yes, but it'd require rewriting a lot of code.
<cizra>
Lears: That is not the case. If I do that, it'll just disappear. journalctl says this: https://paste.ee/p/rXk1b
<Baughn>
There are some mitigations. They mostly amount to "Use a faster computer", though.
<tobiasBora>
Baughn: I think the lab computer is a decent one (core i5)
<Baughn>
Well, which core i5?
<Baughn>
There have been 9-10 generations of processors all called that, so..
<lejonet>
clever: ah, hardenedLinuxPackagesFor sets features.ia32Emulation = false; so I guess I could make my own kernel package that doesn't do that
<tobiasBora>
Baughn: i5-6600 CPU @ 3.30GHz
<Baughn>
Fairly decent, then.
<Baughn>
And 5 seconds for a nixos-rebuild isn't actually a lot.
<Baughn>
...alas, I'm afraid this is just how it is.
<tobiasBora>
Baughn: 5 minutes you mean?
hoijui has joined #nixos
<Baughn>
...oh
<Baughn>
Hm
knupfer has quit [Remote host closed the connection]
<tobiasBora>
I'd be super happy with 5s ^^
<Baughn>
How much memory does it have?
<sphalerite>
tobiasBora: you can pass --fast but it won't be _that_ much faster :p
<Baughn>
No, this is something wrong with the computer.
<Baughn>
Even my own super-complicated nixos builds take <20 seconds.
linarcx has quit [Quit: WeeChat 2.4]
<Taneb>
"write error: Connection timed out at /nix/store/2c462gnvayhliv8x5za35irvmxfwbb0i-perl5.28.1-Starman-0.4014/lib/perl5/site_perl/5.28.1/Starman/Server.pm line 572, <GEN2008> chunk 1."
<Taneb>
In my breaking nixserve
<Baughn>
tobiasBora: Are you at the computer right now?
<tobiasBora>
Baughn: yes
<Baughn>
Ok. Can you run a command for me? One sec..
<tobiasBora>
sure
<Lears>
cizra: If xmonad is installed properly through the nixos options then it won't try to lookup ghc on the PATH; it will use $NIX_GHC as set set in its wrapper script. So those errors suggest you've installed xmonad in some other way.
<tobiasBora>
sphalerite: yes, but as far as I remember the improvement is not huge
<sphalerite>
tobiasBora: is the storage on the machine traditional HDD?
<Baughn>
Well, let's see what diagnosis info we get.
<sphalerite>
nixos likes doing lots and lots of IO on lots and lots of small files
<tobiasBora>
sphalerite: (the script at the end looks interesting, I may try it also for my rasp, I gave up to add/remove modules manually ^^)
<Baughn>
It does. But that *should* be ok, so long as there's sufficient caching.
<Baughn>
At least, not bad to the point of taking 5 minutes.
<Lears>
cizra: lines 88 through 92 are redundant, you can remove them. Otherwise it looks alright. I'm guessing you accidentally have another xmonad on your PATH that's taking precedence over the one installed through nixos options.
<Baughn>
I used to run with /nix on a ZFS HDD pool, for my backup server. It was much slower, but only... 2-3x, not 20x.
<Baughn>
Granted, that's in part because ZFS is awesome.
<tobiasBora>
Baughn: I will try it in 5mn, when lspci is installed :P
<clever>
Baughn: i have 3 machines that are pure zfs
<Baughn>
clever: Oh, all of mine are.
<Baughn>
I meant as opposed to ZFS-on-NVMe
<lejonet>
Baughn: you could swap cat /proc/cpuinfo with lscpu for neater output :)
<clever>
Baughn: the desktop and laptop are zfs on nvme with no redundancy
<Baughn>
lejonet: Neat, but I specifically want the details.
<tobiasBora>
sphalerite: and it's not possible to cache it in RAM or anything? (ok, I don't have lot's of available RAM because nix takes already 20%)
<clever>
Baughn: the nas is zfs on 3 mechanical drives
<Baughn>
tobiasBora: It is.
<clever>
all 3 have the root and nix store on zfs
<Baughn>
tobiasBora: If you run said command, it'll give me some idea of what to tune.
<Baughn>
Oh, and can you throw in `mount` there?
<tobiasBora>
Baughn: I'm running ZFS also if it helps
<lejonet>
Baughn: which ones? lscpu gives you all the details, but without duplicating it for each thread (basically it parses /proc/cpuinfo and prints a summary)
<Baughn>
tobiasBora: And `zfs list`.
cfricke has quit [Quit: WeeChat 2.4]
<Baughn>
tobiasBora: ...and `zfs get all <pool-name>`, as well as `zpool get all <pool-name>`?
<Baughn>
lejonet: The actual current frequency of the cores.
<Baughn>
lejonet: Have seen some cases where the CPU just refuses to go out of minimum power mode.
<lejonet>
Baughn: I see, I don't know how lscpu decides which current freq it should show, but it shows a current freq too
* tobiasBora
is doing tests
<cizra>
Lears: You're right, I have the self-compiled xmonad, compiled with jomik's script. I found how to uninstall packages in manual, but I didn't find how to list them. Or I could utterly nuke any per-user nix-env customizations, how do I do that?
<lejonet>
But yeah, that will most likely mask if just one core/thread is stuck in low power mode
<Philonous>
I'm looking through people's nix files and I'm finding functions like for example haskellPackages.shellFor. What's the best way to find the documentation for it?
<Baughn>
Philonous: If it isn't in the manual, then... er, a google search.
<Baughn>
And/or source code.
<tobiasBora>
It's funny, nixos-rebuild test is faster than nixos-rebuild test --fast ^^ What's the difference
<Baughn>
--fast skips some of the work.
<Baughn>
So that's probably due to cache warmup or whatever.
<cizra>
Lears: Ah, found by googling. nix-env -e '*'
<Baughn>
And I would really like that debug output please? XD
vidbina has quit [Ping timeout: 244 seconds]
<Philonous>
I've tried that. I'm finding github issues, reddit comments, tweets, blog entries, all tangentially related, but no actual documentation I've tried the code search on github without luck
<tobiasBora>
and actually now it's quite faster, don't know why
<Baughn>
I mean, I have a 750GB NVMe in my laptop.
<Baughn>
*768GiB actually.
drakonis has joined #nixos
<tobiasBora>
it's my work computer, so I guess they don't want to spend too much money ^^
<Baughn>
I don't think they build HDDs that small anymore. :P
<Baughn>
But if it's a few years old, then that makes sense.
<Lears>
cizra: Good-o.
<lejonet>
Baughn: judging from it being a 6th gen Intel CPU, its probably an older Dell Optiplex with minimum specs at the time :P
<Baughn>
No obvious tells of a broken HDD. Can't say for sure without running iozone or something, but it seems fine.
<tobiasBora>
SSDs are getting much cheaper now
<tobiasBora>
it's impressive
<lejonet>
Indeed
<tilpner>
tobiasBora: Also check arc_summary, I guess
<Baughn>
Temperature is nominal, no reallocated sectors, reported throughput is normal, no CRC errors either, not dramatically old.
<Baughn>
Also, whoops, I should've used "/dev/sd?".
<__monty__>
Philonous: Clone nixpkgs and grep. Haven't found a better way yet : /
pshevtsov has quit [Quit: WeeChat 2.4]
<Baughn>
tobiasBora: (zfs list; zpool list; zpool get all; zfs get all; arc_summary.pl) | curl -F 'f:1=<-' ix.io
<Philonous>
Nix seems to be suffering from the Haskell illnes of ever-growing layes of abstraction and cobbled together "helper functions" that someone dumps on github and calls it a day. And it doesn't even have the type system to keep in on the rails :(
<Baughn>
Er
<Baughn>
*arc_summary.py
<Baughn>
Philonous: Aye. Frankly, we ought to replace it with TypeScript or something.
<Baughn>
(Not my preferred choice, but it'd work.)
<lejonet>
__monty__: that is usually the only way to get the real details on many things people use, is my experience :/ but, then you get EXACTLY what they do, because you get the code :P
drakonis_ has quit [Ping timeout: 252 seconds]
hoijui has quit [Ping timeout: 245 seconds]
<Baughn>
lejonet: It has 8GiB of memory. So not _minimum_ specs.
<__monty__>
lejonet: Yeah, but I'm not often interested in the details. Just a list of arguments a function supports and preferably what they mean would be great.
<Baughn>
tobiasBora: sudo 'zfs set compression=on sync=disabled zroot'
<Baughn>
tobiasBora: That should help a bit, though be aware that setting `sync=disabled` means fsync becomes a no-op.
<tobiasBora>
Baughn: disabling sync help?
<Profpatsch>
pie__: I think the nix manual
<Baughn>
A lot.
justanotheruser has joined #nixos
<Baughn>
This is 'safe', in that ZFS _never reorders writes across transactions_, hence the observable state after a crash will always be consistent.
<tobiasBora>
I'm curious, why?
<pie__>
Profpatsch, i tried a ctrl-f but didnt find it
<lejonet>
It doesn't wait to hear back from the disk for the writes iirc
<Baughn>
Because with sync=default, an fsync requires multiple _immediate_ writes to disk. And
<Baughn>
*And this is slow.
<Baughn>
That said, beware.
<Baughn>
Disabling sync means fsync does nothing.
<cizra>
I installed autojump. What's the right way to source autojump.bash? It apparently lives in /nix/store/sync2kg7x9a8vgwp3kqzxcn5h82wqjir-autojump-22.5.1 and the nix-build cmd to find that path takes 200ms to execute.
m0rphism has quit [Ping timeout: 246 seconds]
<Baughn>
Which means if the computer crashes, then time will be rolled back a minute or two.
<Baughn>
This can be a proble if it's a server of any kind.
<tobiasBora>
Ok command tried thanks.
<Baughn>
The safer option is to add an slog device, but that would mean adding an ssd.
<Baughn>
As for enabling compression, your CPU is much faster than your disk. Compression equals reading/writing less data.
<Baughn>
Doing that won't have any immediate effect, but it should help once data gets rewritten.
<tobiasBora>
interesting
phreedom has quit [Remote host closed the connection]
phreedom has joined #nixos
<Baughn>
(Plus, small disk. Extra space is extra space. :P)
<tobiasBora>
this command should do something directly or I need to reboot?
<tobiasBora>
(for sync)
<Baughn>
Works immediately.
<Baughn>
Mind you, it only makes _writes_ faster.
<tobiasBora>
ok thank you. Can I force zfs to recompress stuff? ^^
<pie__>
man theres a lot of interesting random stuff in here "Builders can now communicate what build phase they are in by writing messages to the file descriptor specified in NIX_LOG_FD. The current phase is shown by the nix progress indicator. !
<pie__>
s/!/"/
<pie__>
re: __structuredAttrs: " Nix writes a script named .attrs.sh to the builder’s directory that initialises shell variables corresponding to all attributes that are representable in Bash. This includes non-nested (associative) arrays. " i wonder if there's any indication of failure to serialize things
erictapen has quit [Ping timeout: 258 seconds]
<pie__>
"The sandbox-paths configuration option can now specify optional paths by appending a ?, e.g. /dev/nvidiactl? will bind-mount /dev/nvidiactl only if it exists."
xkapastel has quit [Quit: Connection closed for inactivity]
<tobiasBora>
Baughn: I can stop it in the middle?
ottidmes has joined #nixos
<tobiasBora>
Now the time is quite good (I don't if caching is the only reason), I have 8 seconds for a build without changes
<Profpatsch>
pie__: And that’s just the stuff that’s in the changelog!
<Profpatsch>
There’s lots and lots of builtins that don’t even appear in the changelog.
<tobiasBora>
Baughn: do you have other tricks to speed it up even more? ^^
<Baughn>
tobiasBora: If you go the "Copy all your data off the disk, erase the disk, copy it back" route, then you _really_ shouldn't stop in the middle. :P
<Baughn>
tobiasBora: Yeah, replace the HDD with an SSD.
<Baughn>
But this would already be 10-100x faster than your HDD.
<Synthetica>
I can't help but get surprised how cheap ssd's have gotten
<lejonet>
indeed
<infinisil>
And here I am with my almost dying HDD's, too lazy to replace them
<tobiasBora>
that's cheap. But first that's not my computer, second there is a locker on it, third, my laptop has one (very small, like 20Gb :P), but I'm not sure I can simply replace it with another one (actually that would be interesting!)
<pie__>
Profpatsch, /me flails helplessly
<tobiasBora>
Baughn: also, I'm thinking about something else
<tobiasBora>
the processor is supposed to be pretty good, but I think the computer is quite slow
<Baughn>
tobiasBora: SSDs can generally be replaced quite easily. They're just 2.5" form factor SATA devices.
<Baughn>
Didn't spot anything obvious there, but maybe.
<Baughn>
Could be dust.
<tilpner>
tobiasBora: Storage makes a lot of difference in how fast a device feels to use
<tobiasBora>
Baughn: how can you check if the policy is set to performance? I tried "cpupower frequency-info
<tobiasBora>
but it says current CPU frequency: Unable to call hardware
<tilpner>
It doesn't matter if your CPU is fast if all it does is quickly waiting on storage
<tobiasBora>
tilpner: indeed
<tilpner>
Mine says the same, and it's fine here
<tobiasBora>
but I'm curious to check if the governor is set correctly
<tilpner>
I get the same output on laptop and server
<tobiasBora>
ok, so I've no way to check my line "powerManagement.cpuFreqGovernor = "performance";" worked? ^^ (I'm dubious now because my raspberry pi have this exact same problem, cpu freq cannot by modified and I discovered it a bit late)
<tilpner>
*you the
<Baughn>
tobiasBora: The policy shouldn't be set to performance. The default on modern systems is 'powersave', for good reason.
o1lo01ol1o has joined #nixos
<tobiasBora>
ok
<tilpner>
Baughn: It makes some difference for latency on servers
<tobiasBora>
powersave move to performance when needed?
<Baughn>
Yes, but this isn't one.
<Baughn>
It takes the CPU only a millisecond to spin up.
<Baughn>
'powersave' keeps the CPU at low power when idle, and bursts to as high as possible when used, yes.
<tobiasBora>
ok, I'll move it back to powersave then
<Baughn>
It may actually work better. The performance governor depends on having fully functional cooling.
<cizra>
tobiasBora: Plus, as your CPU has thermal mass, it takes a while to heat up. If it runs very cool by default, it'll have a longer opportunity to calculate stuff on Turboboost, before overheating and throttling back down.
<tobiasBora>
ok interesting. I was thinking that powersave was always low power
abathur has joined #nixos
<tobiasBora>
Thank you for all the advices.
<Yaniel>
a cooler CPU also uses less power AFAIK
<tobiasBora>
good for nature, perfect :P
<cizra>
tobiasBora: I noticed it in some games with broken vsync setting - when I was running it on uncapped FPS, things got overheated real fast and my FPS suffered lag spikes. After I fixed frame rate (NVidia's an ass) and installed thermald to use frequency scaling, not fans, for cooling, things went to buttery smooth.
<Yaniel>
because the electrical properties change with temperature (extreme example: superconductors)
<tobiasBora>
Also, one more question. I think people mentionned that it was possible to cache nix stuff in RAM to have quicker access. Any references/quickstart?
leopard has joined #nixos
<cizra>
Isn't just having enough RAM enough? OS should cache it after 1st access.
<Yaniel>
linux does that already behind the scenes
myskran has joined #nixos
<cizra>
How do you guys version your configuration.nix? I don't want to run Git as root, I don't want to keep copying it to /home/user either
abathur has quit [Read error: Connection reset by peer]
<Yaniel>
I don't
astal has quit [Ping timeout: 244 seconds]
<infinisil>
cizra: You can change the location of it
pshevtsov has joined #nixos
<Yaniel>
at my previous workplace we used nixos for a couple of machines, kept their configuration in git and updated them remotely with nixops
<blackriversoftwa>
Anyone have a clue why the fixup phase for compiler-rt moves everything out of the proper store path and into a completely different store path?
<cizra>
infinisil: Heh, I just started importing conf from $HOME in the same way hardware-configuration is imported.
<infinisil>
Oh actually yeah the xz error I've seen too before
<clever>
infinisil: sounds like the tcp connection failed mid way, and that EOF'd xz, which threw its own error
<infinisil>
I didn't remember there being HTTP error 200, which just sounds odd
<gchristensen>
clever's theory sounds right
<clever>
infinisil: i think it 200'd, then began downloading, then died mid-download
<infinisil>
Ah that makes sense
<blackriversoftwa>
Ericson2314: any idea why the fixup phase for a cross-compiling `compiler-rt` would empty out the output store path to a completely different store path and then fail?
<{^_^}>
[nixpkgs] @mkgvt opened pull request #61494 → pytricia: init at version 2019-01-16 → https://git.io/fjlfq
<ottidmes>
I have the situation where I want to base the NixOS configuration of a NixOS installer on another NixOS configuration (the target). To complicate matters, I have several conceptually standalone modules that want to add configuration to the installer config based on the target config, so I was thinking of creating an config option that would be a list of functions going from targetConfig to installerConfig
<{^_^}>
[nixpkgs] @etu opened pull request #61495 → adafruit-ampy: init at 1.0.7 → https://git.io/fjlfs
puck has quit [Remote host closed the connection]
puck has joined #nixos
<Ericson2314>
blackriversoftwa: Not sure. I'd compare the compiler-rt derivation with later LLVM's, which I know work for cross
<Ericson2314>
we really got to deduplicate those!
<blackriversoftwa>
Ericson2314: It seems like it is only installing the include/ directory, and then moving that to the -dev output. But since there is nothing else, it deletes the main store outputs. I'll take a look at the later llvms
<cransom>
we had a running gag at a job where the versioning scheme was a binary counter with old/new since there were multiple machines named/tagged blah-old-new
<cransom>
you could totally do something similar and up with a prority system for prePrePrePostSomething
<pie__>
yes I now have code that says chainWrappers [(debugWrap debug) (prePostWrap "Build")]
tsmeets has quit [Quit: leaving]
knupfer has joined #nixos
<pie__>
cransom, lol i hate^W love you guys
o1lo01ol1o has joined #nixos
<pie__>
if you read the code the prepost actually means running the pre hook and post hooks
<cransom>
minor detail.
drakonis_ has joined #nixos
drakonis_ has quit [Read error: Connection reset by peer]
drakonis_ has joined #nixos
drakonis has quit [Ping timeout: 259 seconds]
InFlames has quit [Ping timeout: 252 seconds]
<{^_^}>
[nixpkgs] @ldesgoui closed pull request #42839 → qutebrowser: shorten launch time marginally if an instance is already running → https://git.io/fwl0s
mkoenig has quit [Ping timeout: 244 seconds]
Glider_IRC_ has joined #nixos
mkoenig has joined #nixos
vidbina has joined #nixos
primeos has quit [Quit: WeeChat 2.4]
_e has quit [Quit: WeeChat 2.4]
civodul has joined #nixos
noam has joined #nixos
Glider_IRC has quit [Ping timeout: 258 seconds]
primeos has joined #nixos
<pie__>
is it possible for nix to keep a build that failed its hash check?
<pie__>
basically i want a "here build this thing" "oh no it failed" "yeah well here have this thing with the hash i just corrected"
drakonis has joined #nixos
<clever>
pie__: there is a prefetch flag that will do that
<pie__>
im probably doing Bad Things but...we'll see
noam_ has quit [Ping timeout: 268 seconds]
<pie__>
clever, any idea how to find that?
<clever>
pie__: basically, if a fixed output derivation has the wrong hash, it renames the $out to match its hash
<clever>
that means it will keep whatever the artifact is, and the path its at will match its hash
<pie__>
im guessing that matches the way nix looks for stuff
<samueldr>
yeah, you can observe it in a TOFU situation, for a slow download
<pie__>
TOFU?
<samueldr>
at least, in my case
<samueldr>
,tofu
<{^_^}>
To get a sha256 hash of a new source, you can use the Trust On First Use model: use probably-wrong hash (for example: 0000000000000000000000000000000000000000000000000000) then replace it with the correct hash Nix expected.
<clever>
but its not the path your fetchurl was giving
drakonis_ has quit [Ping timeout: 252 seconds]
<clever>
if you change the fetchurl to use the right hash, it will then discover the artifact it previously downloaded
<clever>
and instantly use it
<pie__>
samueldr, ah right
<pie__>
clever, that makes sense
<pie__>
i think ive run into that a lot more with a recent update
<pie__>
(recent?)
jasongrossman has quit [Ping timeout: 268 seconds]
erictapen has joined #nixos
<pie__>
because basically im using a fixed output derivation to hack myself into having network acces
<pie__>
^i mean context for the question
<pie__>
and i want to be able to store its output without user intervention
_e has joined #nixos
<Taneb>
...Doesn't nix cache all fetches for some time?
<{^_^}>
[nixpkgs] @vbgl opened pull request #61499 → coq: init at 8.10+β1 → https://git.io/fjlUE
<pie__>
clever, maybe i shouldnt be using nix as basically a database lol
sepi has quit [Ping timeout: 246 seconds]
jabranham has joined #nixos
hoijui has joined #nixos
_e has quit [Ping timeout: 250 seconds]
hoijui has quit [Remote host closed the connection]
<blackriversoftwa>
Ericson2314: patching the LLVM7 cmakeFlags stuff for compiler-rt back to LLVM6 seems to have done the trick :)
<blackriversoftwa>
Ericson2314: thank you
asheshambasta_m has joined #nixos
srid has quit [Ping timeout: 246 seconds]
dbmikus has joined #nixos
<yorick>
oi! nix-prefetch-url no longer works in a nix build?
<yorick>
I just need to hash some files
<gchristensen>
did it ever? what did it do?
<yorick>
hash some files, I suppose
<yorick>
well, it's here in this buildPhase
<sphalerite>
might work outside the sandbox
Jackneill has quit [Ping timeout: 248 seconds]
<pie__>
yorick, i patched nix-prefetch-git locally yesterday so i could use go2nix in the sandbox
<pie__>
it was using it just to get the has...
<pie__>
*hash
<yorick>
what's the nix-hash equivalent to nix-prefetch-url --sha256 --unpack file://asdf?
<pie__>
(maybe implicitly "pre"downloading the stuff into the store i guess)
<pie__>
idk i just commented out the nix-store --add line
astrofog has quit [Remote host closed the connection]
<clever>
the issue, is that systemd.nix adds some default things to PATH, so the normal Environment.PATH cant be set, since the whole Enrivonment key lacks merge rules
<clever>
so a dedicated option exists for path, that knows how to merge
<samrose>
clever: ok that makes sense, and explains why you cannot just set it in the `environment` for a systemd service
<{^_^}>
[nixpkgs] @Lassulus pushed commit from @bhipple to staging « pkg-config: enable check »: https://git.io/fjlIl
dbmikus has joined #nixos
pie_ has joined #nixos
ggpeti has quit [Ping timeout: 256 seconds]
galaxie has left #nixos [#nixos]
<jluttine>
I cannot get Kwallet to auto-unlock on login. I'm not using KDE, but I've set security.pam.services.sddm.enableKwallet=true. What else do I need to do?
<jluttine>
(I'm using sddm as my display manager)
<jluttine>
Or does kwallet even provide org.freedesktop.secrets, or is gnome keyring the only service to provide that?
asheshambasta_m has quit [Quit: Leaving.]
<jluttine>
Kwallet did auto-unlock on plasma, so there's some setup that it does differently but I can't figure out what
o1lo01ol1o has quit [Remote host closed the connection]
pshevtsov has quit [Ping timeout: 255 seconds]
o1lo01ol1o has joined #nixos
pshevtsov has joined #nixos
alexherbo2 has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
T_S_ has joined #nixos
T_S_ has quit [Client Quit]
erictapen has quit [Ping timeout: 268 seconds]
knupfer has quit [Ping timeout: 250 seconds]
o1lo01ol1o has joined #nixos
lovesegfault has quit [Ping timeout: 258 seconds]
jasongrossman has joined #nixos
asheshambasta_m has joined #nixos
Myrl-saki is now known as Miyu-saki
leopard has quit [Ping timeout: 246 seconds]
knupfer has joined #nixos
freeman42x has joined #nixos
fresheyeball has joined #nixos
erictapen has joined #nixos
lordcirth_ has quit [Remote host closed the connection]
lordcirth_ has joined #nixos
orivej has joined #nixos
<fresheyeball>
sh: binwrap-install: command not found
<fresheyeball>
anyone seen this when trying to install a nodePackage?
<fresheyeball>
It would appear `nixpkgs.nodePackages.elm-test` is broken
callahad has quit [Quit: Ping timeout (120 seconds)]
o1lo01ol1o has joined #nixos
callahad has joined #nixos
toppler`` has quit [Ping timeout: 246 seconds]
o1lo01ol1o has quit [Remote host closed the connection]
toppler`` has joined #nixos
esmerelda has quit [Ping timeout: 245 seconds]
<blackriversoftwa>
Ericson2314: Actually it looks like I mistook the llvm targeting my build machine for the one targeting ARM. That one fails to apply the crt_begin_and_end patch to LLVM 6.
<blackriversoftwa>
I'm not quite sure how to backport it
Myhlamaeus has joined #nixos
esmerelda has joined #nixos
MmeQuignon has quit [Ping timeout: 252 seconds]
o1lo01ol1o has joined #nixos
<Ericson2314>
blackriversoftwa: maybe just do the patch without it (it still has some use) and use llvm 7
domogled has joined #nixos
<blackriversoftwa>
Ericson2314: I'm trying to use llvm 6 because I'm using ghc-8.6.4 as a cross-compiler (via the IOHK haskell.nix infrastructure) and keep hitting weird errors with iserv
<blackriversoftwa>
So I want to use a stdenv where llvm 6 clang is the c compiler, to see if that helps, instead of mixing llvm for haskell code and gcc for c code in Haskell packages.
<Ericson2314>
iserv for template haskell?
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
katyucha has joined #nixos
<katyucha>
Hi !
o1lo01ol1o has quit [Remote host closed the connection]
mmlb416 has quit [Quit: Ping timeout (120 seconds)]
hellrazor has joined #nixos
mmlb416 has joined #nixos
<katyucha>
I try a rebuild to modify a service... and I have this error : You have a memory leak (not released memory pool):
<blackriversoftwa>
Ericson2134: It's a brilliant hack for cross-compilation with template haskell
talqu has joined #nixos
o1lo01ol1o has joined #nixos
<blackriversoftwa>
It lets you build an executable to run template haskell splices on the target platform, while letting a build-platform ghc call out to that executable
<blackriversoftwa>
so you can run your template haskell splices under something like qemu or wine, while ghc runs natively on your build machine
<blackriversoftwa>
But in my case it keeps running up against linker errors on 32-bit arm
<blackriversoftwa>
I've been hammering on this for over a week with angerman, and he suggested I ask you about the LLVM stuff.
acarrico has quit [Ping timeout: 245 seconds]
sigmundv has joined #nixos
<Ericson2314>
blackriversoftwa: so building the thing to run with iserv?
o1lo01ol1o has quit [Ping timeout: 244 seconds]
<Ericson2314>
or nevermind
<Ericson2314>
anyways so I haven't used iserv much
noam has quit [Ping timeout: 258 seconds]
<Ericson2314>
do you us the binfmt thing?
<Ericson2314>
or have a impure derivation?
<blackriversoftwa>
neither
<blackriversoftwa>
the derivation has qemu as a build dependency
<Ericson2314>
ok
<blackriversoftwa>
and passes a flag to GHC telling it to run an external interpreter under qemu
<blackriversoftwa>
but that isn't the issue
<Ericson2314>
yeah I've never tried iserv
johnw has quit [Remote host closed the connection]
<Ericson2314>
we at Obsidian have a patch for GHC that dumps splices
o1lo01ol1o has joined #nixos
<blackriversoftwa>
iserv is just ghci but for ghc to talk to over a pipe
mexisme has joined #nixos
<blackriversoftwa>
so if you run iserv under qemu, then your TH splices run under qemu
<Ericson2314>
so we do a native build and dump the TH and then load it in the cross build
<blackriversoftwa>
Yeah I've seen obsidian's stuff
<Ericson2314>
what sort of linker errors do you get?
<Ericson2314>
you should in principle be able to use libgcc to sidestep the missing stuff in compiler-rt
johanot has joined #nixos
<blackriversoftwa>
Ericson2314: `Failed to lookup symbol: _GLOBAL_OFFSET_TABLE_`
<blackriversoftwa>
That's when I just build with a regular cross-compiling stdenv
<blackriversoftwa>
I'd like to use an LLVM 6 stdenv
<Ericson2314>
well you'll need crtbegin and end somehow
wfranzini has joined #nixos
<Ericson2314>
and forever llvm people were hagling over who ought to provide it
<blackriversoftwa>
great
o1lo01ol1o has quit [Ping timeout: 268 seconds]
<Ericson2314>
cause it comes with os-specific libc on freebsd darwin etc
<Ericson2314>
that's why it wasn't in it from the get-go
<Ericson2314>
but if you want 100% no GCC stdenv
<Ericson2314>
*no GCC or glibc
<Ericson2314>
then you need to back-port that thing somehow
waleee has quit [Quit: WeeChat 2.4]
<blackriversoftwa>
Well I'm just operating on a theory (not my own) that the linker error had to do with the Haskell code being built with LLVM, but the included c bits being built with GCC
<Ericson2314>
_GLOBAL_OFFSET_TABLE_ could have something to do with -fpic and -fpie
<blackriversoftwa>
Ericson2314 believe me I've been trying to chase that down
lovesegfault has quit [Ping timeout: 244 seconds]
o1lo01ol1o has joined #nixos
<blackriversoftwa>
Ericson2314: Is that error because something was built with -fPIC and shouldn't have been?
<blackriversoftwa>
Or wasn't built with -fPIC and should have been?
<blackriversoftwa>
That has been confusing me for a week now
<Ericson2314>
I would guess that more position indepedence <=> more indirection <=> more _GLOBAL_OFFSET_TABLE_
<Ericson2314>
and generally that is what you want
kreisys has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Ericson2314>
are you doing iOS / android builds? those definitely need position independent code
<blackriversoftwa>
no I'm building stuff for Raspberry Pi
<blackriversoftwa>
but unfortunately these Pis are in the middle of nowhere running 32 bit OSes
<Ericson2314>
that's cool though :D
<blackriversoftwa>
otherwise I'd just build for Aarch64 and have done with it
<Ericson2314>
but yeah I feel ya
fendor has quit [Ping timeout: 245 seconds]
<Ericson2314>
no idea whether GOT/pit is the expectation
<blackriversoftwa>
Armv7l seems to be the ugly duckling of linux targets anymore
myskran has quit [Ping timeout: 245 seconds]
<Ericson2314>
yeah all the 32-bits are rotting a bit
<Ericson2314>
it seems
run500 has joined #nixos
<Ericson2314>
it's silly, redhat gottta learn to CI or whatever
freeman42x has quit [Ping timeout: 258 seconds]
<blackriversoftwa>
Ericson2314: I've made sure GHC was being called with pic and pie. Everything works until iserv tries to load a package that was built with some linked c code (cryptonite in my case)
<run500>
im curious why git fetchers require a sha256 hash. doesnt the git rev hash already hash the content?
<Ericson2314>
fucking cyptinite...
<Ericson2314>
it's always some shit like that
<blackriversoftwa>
For a long time I couldn't get anything that needed iserv to build, but then angerman suggested adding pic and pie to the "hardeningDisable" flags
<Ericson2314>
so it *wasn't* adding those
<blackriversoftwa>
Ericson2314: Have you had issues with cryptonite before?
<Ericson2314>
hmmm
zupo has joined #nixos
<blackriversoftwa>
Well before I added those even Aeson wouldn't build
<Ericson2314>
me not personally but I recall some in static builds, and some in a client with vaious CIs
<Ericson2314>
symbol clobbers with multiple crypto libraries
<blackriversoftwa>
Ericson2314: any leads on this?
<Ericson2314>
all the haskell libraries seem to just cut and paste C in sloppy ways
<blackriversoftwa>
Ericson2314: If someone is a Haskell programmer it is often because they don't want to be a C programmer ;)
<Ericson2314>
those weren't like this, but this seems like the hardening disable meant not enough things got pic and pie
<blackriversoftwa>
(Certainly that is true in my case)
logzet_ has quit [Remote host closed the connection]
polman has joined #nixos
<gchristensen>
you want an epub or a pdf?
o1lo01ol1o has quit [Remote host closed the connection]
<devalot>
Networking question: my laptop will either be connected to a wired network or wireless depending on location. I want my containers and VMs to be behind NAT. Right now I have nat.externalInterface set to my wireless interface. But of course that breaks when I'm connected via a wired connection.
<__monty__>
It can generate pdf too of course. Just don't know if that comes with a useful toc.
IRCsum has quit [Ping timeout: 250 seconds]
<devalot>
I thought maybe I wanted to bridge my two NICs. But I don't think that's right.
clever_ has joined #nixos
<tolt>
gchristensen: ultimately I need a pdf (using a sony digital paper, only format it reads) but I can convert an epub
vid has quit [Ping timeout: 258 seconds]
<__monty__>
tolt: I'd still try pandoc. You can probably generate a pdf with a toc.
clever has quit [Ping timeout: 245 seconds]
clever_ has joined #nixos
clever_ has quit [Changing host]
clever_ is now known as clever
IRCsum has joined #nixos
<{^_^}>
[nixpkgs] @c0bw3b opened pull request #61512 → winstone: drop package and service → https://git.io/fjlmm
<{^_^}>
[nixpkgs] @zimbatm opened pull request #61513 → WIP: cryptpad: init at v2.21.0 → https://git.io/fjlmY
<Qubasa>
devalot: So know it starts automatically after every rebuild with: the following new units were started: sys-devices-virtual-net-wireguard\x2dhome.device, sys-subsystem-net-devices-wireguard\x2dhome.device
talqu has quit [Remote host closed the connection]
work_ has quit [Quit: Connection closed for inactivity]
o1lo01ol1o has joined #nixos
<nh2>
matthewbauer: for my build there sandboxing is enabled. But it says "unshare: command not found" there
<nh2>
matthewbauer `unshare` is usually in utillinux, but that doesn't seem to be a declared nix dependency
<gchristensen>
wireguardians, do y'all see modprobe: FATAL: Module wireguard not found in directory /run/booted-system/kernel-modules/lib/modules/... when deploying a system with a kernel update?
o1lo01ol1o has quit [Remote host closed the connection]
<andi->
haven't seen that on my wireguard machines. Not entirely sure how the kernel update timing is. I usually reboot those as soon as a new kernel is deployed.
o1lo01ol1o has joined #nixos
<{^_^}>
[nixpkgs] @flokli pushed commit from @ambrop72 to release-19.03 « kernel: Disable VirtualBox guest modules. »: https://git.io/fjlOc
adamantium has joined #nixos
<gchristensen>
oh ... right, it is trying to load the modul ebecause it never succeeded to because it doesn' thave the module available.
lovesegfault has joined #nixos
jackdk_ has quit [Ping timeout: 246 seconds]
alexherbo2 has joined #nixos
o1lo01ol1o has quit [Ping timeout: 258 seconds]
callahad has quit [Quit: Ping timeout (120 seconds)]
callahad has joined #nixos
esmerelda has quit [Ping timeout: 245 seconds]
esmerelda has joined #nixos
<{^_^}>
[nixpkgs] @adelbertc opened pull request #61521 → buildRustPackage: fix cargo flag for release/debug build → https://git.io/fjlOa
fendor_ has quit [Ping timeout: 255 seconds]
<colemickens>
Is there something concrete I can do to help the mesa/libdrm meson changes in or into staging? I'm not really sure if there's something we're waiting on or who owns that space, or how big things like mesa upgrades are usually handled.
<{^_^}>
[nixpkgs] @tadeokondrak opened pull request #61522 → vapoursynth-editor: init at R19 → https://git.io/fjlOM
<zeta>
hello guys, i am testing my irc, can you guys see this message?
<gchristensen>
I guess I'd like to see a review from jtojnar colemickens
jluttine has quit [Ping timeout: 255 seconds]
mbrgm_ has joined #nixos
<zeta>
testing, testing, did you guys get it?
<gchristensen>
yep
mbrgm has quit [Ping timeout: 252 seconds]
mbrgm_ is now known as mbrgm
<worldofpeace>
if I can find time i can give you a review colemickens
<worldofpeace>
we must meson the world! :P
<colemickens>
okay, thanks. I don't mean to come across impatient, I'm actually more interested in the mechanics of how "we" try to stage such changes to staging, and then eventually to master.
<ivan>
does anyone know which kwin thing is responsible for creating a 1px-wide window on the left edge of the screen that blocks scrolling?
<gchristensen>
colemickens: things go in to staging at will, and then periodically staging-next is branched from staging, stabilized, and merged tomaster
kisonecat has joined #nixos
<colemickens>
presumably a similar set of people is going to be reviewing the libdrm/mesa/mesa-related PRs and trying to get the mass rebuilds done together.
<v0id72[m]>
was wondering if someone could help me with an issue I am having in 19.03. For some reason an external drive keeps dropping connection and not letting me write to it.
<v0id72[m]>
i have it mounting but not able to write.
<v0id72[m]>
As I wrote the info in the wiki about the legality issue to exFat
bourbon has joined #nixos
<bourbon>
my coworker said "since we're already using nix, just install it with nix"
<bourbon>
here I am, six hours later still no closer to adding telegraf through nix
<bourbon>
how does one tell nix it should install a derivation in your home.nix or default.nix or whatever.nix