gchristensen changed the topic of #nixos-chat to: NixOS but much less topical || https://logs.nix.samueldr.com/nixos-chat
<drakonis> aw
ldlework has quit [Remote host closed the connection]
ldlework has joined #nixos-chat
<abathur> heh
julm has quit [Remote host closed the connection]
julm has joined #nixos-chat
Emantor has quit [Quit: ZNC - http://znc.in]
Emantor has joined #nixos-chat
rajivr has joined #nixos-chat
cjpbirkb1 has joined #nixos-chat
cjpbirkbeck has quit [Ping timeout: 264 seconds]
<lovesegfault> jtojnar: when you're able to, could you give #98917 a last look?
<{^_^}> https://github.com/NixOS/nixpkgs/pull/98917 (by lovesegfault, 1 week ago, open): klipper: init at 0.8.0
slack1256 has joined #nixos-chat
waleee-cl has quit [Quit: Connection closed for inactivity]
slack1256 has quit [Remote host closed the connection]
endformationage has quit [Ping timeout: 240 seconds]
<samueldr> ooh... less than a week up to nixcon
<samueldr> for the hackday, I might be open to remote pair port with someone that needs it for Mobile NixOS
cole-h has joined #nixos-chat
<cole-h> Hello my friends
<cole-h> I am typing this from my brand new AMD system
<samueldr> hi
<samueldr> how's the buyer's remorse?
<cole-h> None at all tbh
<cole-h> A little sad I missed out on some RAM savings by being impatient
<cole-h> But overall happy
<lovesegfault> cole-h: Nice!
<etu> cole-h: Very nice :)
<cole-h> 8c/16t feels good. So does 32G of RAM :D
<cole-h> Now I gotta fix libvirt.
cole-h has quit [Quit: Goodbye]
cole-h has joined #nixos-chat
cole-h has quit [Client Quit]
<lovesegfault> goddamn the cold is making my right foot hurt so much
cole-h has joined #nixos-chat
cjpbirkb1 has quit [Quit: Goodbye, take care]
<cole-h> Anybody ever run into "No irq handler for vector" on an AMD chip? (I have a 3700x, specifically)
<cole-h> (Kernel output on boot)
<eyJhb> samueldr: I have a friend who "lost" 400.000 DKK, if he had just waited another month bying his apartment. He had some!
<eyJhb> cole-h: Wish I had that much ram/CPU/stuff :(
<ar> i'm still on a 2700x, but with 64GB of ram
<ar> and i plan to build a new box around a 5950x
<eyJhb> ar: do you have a use for all that power?
<sphalerite> eyJhb: is that 400000 or 400?
<eyJhb> THat is
<eyJhb> > DKK 400000
<{^_^}> "1386004000.000000 VND"
<eyJhb> > VND 1386004000
<{^_^}> "52668.152000 EUR"
<sphalerite> ok that's a lot
<eyJhb> Yeah, he bought his apartment like a couple of days before the reevaluated all the apartments in that district...
<sphalerite> in English, . is used as a decimal separator, not for grouping digits :)
<ar> eyJhb: yup. and i wanted to get a 3950x, but between being able to get it and zen3 getting close to release, i decided i can wait a few more months
<eyJhb> sphalerite: not sure if one can specify it as in english that does that, but rather GB and US :p But I get your point
<eyJhb> But 400 DKK would have been OK for him :D
<eyJhb> But it is harsh that the game is like that, screwing somebody over knowing that it will happen..
<sphalerite> eyJhb: ok, but https://en.wikipedia.org/wiki/Decimal_separator#Countries_using_decimal_comma South Africa is the only English-speaking country listed there, unless I missed one
<eyJhb> ar: what do you use all that POWER for
<ar> eyJhb: mostly running VMs of stuff
<eyJhb> sphalerite: now I just hate the 400'00
<sphalerite> eyJhb: I like the General Conference on Weights and Measures declaration though. Either , or . for decimals, and only spaces for digit grouping.
<eyJhb> So - 400 000,00 ?
<sphalerite> yes
<sphalerite> or 400 000.00
<sphalerite> "In 1958, disputes between European and American delegates over the correct representation of the decimal separator nearly stalled the development of the ALGOL computer programming language."
<sphalerite> lol
<eyJhb> sphalerite: sounds about right
<sphalerite> Ugh, I thought I was onto what was causing some weird crashes โ€” the gcc-lib store path was corrupted. But nooooo, it was just __pycache__, no filesystem corruption
cole-h has quit [Ping timeout: 240 seconds]
<eyJhb> sphalerite: saturday project?
<clever> sphalerite: french still uses , as in 3,14 is pi
<clever> ah, only read half the above convo, lol
__monty__ has joined #nixos-chat
MichaelRaskin has quit [Ping timeout: 258 seconds]
vesper11 has quit [Ping timeout: 240 seconds]
nullheroes has joined #nixos-chat
MichaelRaskin has joined #nixos-chat
waleee-cl has joined #nixos-chat
endformationage has joined #nixos-chat
abathur has quit [Quit: abathur]
cole-h has joined #nixos-chat
abathur has joined #nixos-chat
<aleph-> samueldr: So enjoy your helios64 assembly.
<aleph-> Good God does this suck
<__monty__> It comes as a kit?
<aleph-> Eyep
<aleph-> Thought you could get it assembled. Evidently not.
<__monty__> Not much to assemble though, is there?
<__monty__> Isn't it board, chasis, power brick?
<aleph-> Basically. Just cramped as crud
<aleph-> Getting the sata connectors in and the like took ages
<abathur> anyone aware of a generic way to ~feature-detect support for specific /etc/<conf>.d/ drop-in config directories?
cjpbirkbeck has joined #nixos-chat
cole-h has quit [Quit: Goodbye]
cole-h has joined #nixos-chat
<samueldr> aleph-: whenever it gets off the boat, and comes to me, I surely will
rajivr has quit [Quit: Connection closed for inactivity]
<sphalerite> abathur: what do you mean exactly?
<sphalerite> abathur: as in detect that a piece of software is capable of scanning a .d directory for configuration?
<abathur> yeah; some systems (probably more like some *versions* of some systems, or versions of other software on them) support it--but if you can't detect it, the lowest-common-denominator (unified file) still has a lot of appeal/inertia
<colemickens> "9554 store paths deleted, 17995.42 MiB freed" pssh right?
<colemickens> how about "note: currently hard linking saves 17592186044416.00 MiB"
<ldlework> NixOS friends, do any of you know what causes SSHFS mounts to become unresponsive causing it to be unmountable (with device busy) and causing every new shell to hang untill sshfs proc is force-killed?
<MichaelRaskin> I think some kinds of network failures lead to this?
<MichaelRaskin> You can also abort the FUSE connection instead of killing the sshfs process if you prefer
<joepie91> yep, that's common for network failures
<__monty__> colemickens: Is that 17 Exabytes of saved space? ๐Ÿคจ
<MichaelRaskin> Only 18GB
<__monty__> MichaelRaskin: That's space *freed* not saved by hardlinking.
<MichaelRaskin> Oh
<MichaelRaskin> It is very honest and can notice negative savings; an unsigned integer on the output path spoils it a bit, though
<ldlework> MichaelRaskin: abort the FUSE connection?
<ldlework> `fusermount -u` just complains the "device is busy"
<MichaelRaskin> Yep. If you mount the fusectl file system on /sys/fs/fuse/connections, you will have a list of FUSE connections
<MichaelRaskin> You can about one of them if you wish
<ldlework> OK I can google that, thanks.
<MichaelRaskin> Doesn't seem to be any better than killing sshfs in practice
<MichaelRaskin> If you have a single advanced daemon serving multiple FUSE mountpoints with error recovery, it might make more sense
<sphalerite> ldlework: you can also lazy-unmount it with -z, that will at least remove it from the filesystem and prevent any _more_ processes from hanging on it.
<sphalerite> I suppose that's most useful when you still have some expectation of it recovering, though.
<__monty__> MichaelRaskin: Wait, how can hardlinking save negative space? Isn't a hardlink just a directory entry and don't all files have at least one of those?
<samueldr> if there's a hardlink to something that's not in the store I presum
<samueldr> presume*
<samueldr> imagine your store has only one file, a file named A, and your hardlink dir only has one file, a hardlink to a file Z
<MichaelRaskin> __monty__: adding a second hardlink to a file only used once, as Nix indexing does, can increase the size of .links directory without reducing the size of anything
<samueldr> shouldn't that be +1 -1, so zero?
<samueldr> oh, _second_ hardlink
<MichaelRaskin> Well, --optimise _adds_ one .links content-addressed hardlink for everything in store
<MichaelRaskin> They are GC-ed of course, but as long as nothing is actually reused they might be a net loss anyway
<__monty__> Ah. I assumed it'd only start creating hardlinks if it found duplicate files.
<MichaelRaskin> That's hard to do incrementally
<aleph-> samueldr: Yeah will love to hear back about any issues you have
<drakonis> ldlework: hmm what are you trying to do atm
<ldlework> nothing, my sshfs mount keeps hanging is all
<aleph-> Got mine assembled and it seems to do nothin when I have the power cable in except light up >_>
<samueldr> I don't know if it ships with software
<samueldr> so it _could_ be normal
<__monty__> ldlework: I've found sshfs prone to this as well.
<__monty__> NFS is more stable but a lot slower. I've yet to try samba (couldn't be bothered) but many people say it's faster (possibly due to NFS misconfiguration).
<aleph-> samueldr: Yah it doesn't ship, but seems the serial cable isn
<aleph-> isn't even working.*
<samueldr> I couldn't say for now
<aleph-> Hmm, time to dissasemble it I guess.
<aleph-> Aye, true.
<MichaelRaskin> __monty__: then there is XtreemFS and other more exotic options
<__monty__> Haven't heard of it. I'll add a tab to the backlog.
endformationage has quit [Quit: WeeChat 2.9]
endformationage has joined #nixos-chat
<aleph-> Ah that's an issue. samueldr be forewarned, the usb-c (serial cable) port seems to be deeper then the backplate allows when fully assembled. IE: The usb-c prong on their supplied cable doesn't actually fit unless your take off the fans/backplate first annoyingly.
<samueldr> :|
<samueldr> do you have an estimate by how many mm it doesn't fit?
<samueldr> e.g. could you shave it off from the cable?
<samueldr> or would embigenning the plate's hole be easier?
<aleph-> Great question. Maybe honestly a 1mm?
<aleph-> I'll take some pics in a few to show you.
<samueldr> right, so shaving might be the easiest way to go
<aleph-> Honestly annoying that I can't seem to turn off/restart the thing either. Maybe I need to let the ups die off/unplug it and then plug everything in...
cjpbirkb1 has joined #nixos-chat
cjpbirkbeck has quit [Ping timeout: 246 seconds]
<aleph-> And I think I need to dissasemble the whole thing to get the ups unplugged... sigh
<aleph-> Be great if they had a forum
cole-h has quit [Ping timeout: 240 seconds]
<ashkitten> maybe i should set up a system to automatically rebase my nixpkgs onto unstable
<ashkitten> and deploy it to my systems
<MichaelRaskin> What could possibly go wrong. (Said me, while rebuilding my system from a random master checkout)
<ashkitten> hey at least unstable has tests
<ashkitten> i've only had a few things break
<ashkitten> maybe it'd be better to just have a calendar event each week to update my shit
<ashkitten> otherwise i'll never do it
<abathur> did you see what graham was working on a few days ago?
<abathur> (not that it handles rebasing, to avoid getting your hopes up... :)
<ashkitten> no
<abathur> I guess this is the most-succinct log section https://logs.nix.samueldr.com/nixos-chat/2020-10-08#1602159301-1602158815;
kloenk has quit [Ping timeout: 272 seconds]
kloenk has joined #nixos-chat
<eyJhb> It is pretty cool!
<eyJhb> Even if I did not understand a single thing that was happening.
<eyJhb> But generally graham is doing kill stuff with his NixOS configs atm.!
<aleph-> samueldr: Okay yeah totally gonna bother you when you get yours haha
<aleph-> Can't get this thing booting at all
<aleph-> The joy of open, small batch hardware
<aleph-> There we go
<aleph-> Man this thing is super fiddly
<__monty__> Fiddling helped?
<sphalerite> aleph-: oh, no, this bodes ill!
<sphalerite> I guess it'll still be better than my current NAS though.
<aleph-> __monty__: So if you plug power in without an sd card plugged in it'll get stuck in an infinite boot of some kind till any power dies
<aleph-> __monty__: sphalerite: Gonna take pics of the backplate in a bit, super, super fiddly. On my version, with the plate on I can't plug in a mini-sd or a usb-c cable
<aleph-> And the images are super crufty, ssh doesn't even work despite running.
<samueldr> it was likely in rockusb mode
<samueldr> wouldn't reset work?