<manveru>
i really wish nixos-rebuild didn't insist on the flakeAttr being in `nixosConfigurations`...
orivej has joined #nixos-dev
Jackneill has quit [Read error: Connection reset by peer]
Jackneilll has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
orivej_ has joined #nixos-dev
<gchristensen>
Mic92: that was a pretty funny reply to Michael Dexter :'D. I've been whispering about Nix in his ear for a while now.
<FRidh>
using nix dev-shell I get derivation '/nix/store/lxkiicrxvf4fxad3w5ixac4wqwqsqbz7-python3.8-pyacodb-dev-env.drv' has incorrect output '/nix/store/kxlsxgxnp9x49bh8zk6dwi6fdc2xk7lv-python3.8-pyacodb-dev-doc', should be '/nix/store/hbln0va6p7mbi0xa2mdb8nqi2y4a0pi6-python3.8-pyacodb-dev-env-doc'
<FRidh>
flakes bug?
orivej_ has quit [Ping timeout: 258 seconds]
phreedom_ has joined #nixos-dev
orivej has joined #nixos-dev
<FRidh>
using only a single output "out" bypasses the issue
phreedom has quit [Ping timeout: 240 seconds]
raboof has quit [Ping timeout: 260 seconds]
raboof has joined #nixos-dev
<Mic92>
gchristensen: have you used freebsd?
<gchristensen>
no and I don't ever plan on it
<gchristensen>
is it nice? :)
<gchristensen>
(I don't mean to say I'm avoiding it or won't use it, I just have no plan to do so)
<evax>
FRidh: are you using mkShell? if so I've had a similar issue last week and worked around it using symlinkJoin instead
<FRidh>
evax: this is with buildPythonPackage
<FRidh>
evax: you did not have multiple outputs then I suppose?
<Mic92>
gchristensen: I am not super convinced by it. It seemed to be stuck in the past. Their configuration basically consists of shell scripts, the POSIX one without at least some basic data structure. Writing rc.d scripts is not exactly a pleasure. For long time they did not had a proper package manager. They bring up arguments like zfs and dtrace, but now we have in Linux the same or aquivalents
<Mic92>
that are at least as good.
<gchristensen>
I agree
<gchristensen>
my real motivation is to tempt him to try NixOS :P
<{^_^}>
nix#3770 (by FRidh, 45 seconds ago, open): nix develop: derivation ... has incorect output ..., should be ...
<Mic92>
They claim that userspace and kernelspace are better integrated since they are developed by the same team. However the loosley coupling is what made things like docker possible. In FreeBSD jails have to be updated with every sysupgrade or else the userland might not work anymore.
<gchristensen>
oof
<qyliss>
The most interesting thing about FreeBSD for me is Capsicum
<qyliss>
Which doesn't seem to be going anywhere (or an equivalent) on Linux
<qyliss>
it's the least interesting of the BSDs though imo
<qyliss>
Not at all. Capsicum is proper capabilities. Hardening is trying to restrict what a process does one thing at a time and hoping you don't break anything.
<Mic92>
What are the pratical implications?
<qyliss>
It's much for likely for a process to have the minimal capabilities it needs to be able to do its job, and therefore less likely to be able to be exploited
<qyliss>
With Linux/systemd, for example, you can either grant CAP_NET_ADMIN to a service, or you can not. But the surface area of CAP_NET_ADMIN is huge and dangerous, especially considering the relative safety of some operations it's required for.
<Mic92>
I would became more granular. Systemd also has now BPF filter to restrict socket operations further
<LnL>
I had weird issues recently but that had nothing to do with firefox
<LnL>
somehow ended up with an ipv6 only connection and unlike a few years ago many things still worked :D
<timokau[m]>
gchristensen++ that looks great. Its probably just a hand-written design and not actually implemented in any way right?
<{^_^}>
gchristensen's karma got increased to 325
<gchristensen>
yeah I made that in a google doc :P
<timokau[m]>
Ah, that's a shame :D Still a good start, maybe somebody will work on this. I've wanted something like that for a while too.
__monty__ has joined #nixos-dev
<timokau[m]>
At that point you'd probably have to launch a proper TUI to get full control of the "screen" and the ability to redraw not only the last line. Then you could also add the last line of every running build by default, instead of squishing it all into one line.
justanotheruser has quit [Ping timeout: 260 seconds]
drakonis has joined #nixos-dev
justanotheruser has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
ericnoan has quit [*.net *.split]
ryantm_ has quit [*.net *.split]
notgne2 has quit [*.net *.split]
marek has quit [*.net *.split]
evax has quit [*.net *.split]
arianvp has quit [*.net *.split]
ericnoan has joined #nixos-dev
marek has joined #nixos-dev
notgne2 has joined #nixos-dev
ryantm_ has joined #nixos-dev
evax has joined #nixos-dev
justanotheruser has quit [*.net *.split]
xwvvvvwx has quit [*.net *.split]
xwvvvvwx has joined #nixos-dev
justanotheruser has joined #nixos-dev
arianvp has joined #nixos-dev
justanotheruser has quit [Max SendQ exceeded]
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-dev
justanotheruser has joined #nixos-dev
ris has joined #nixos-dev
* colemickens
wants magic tooling to split apart a PR based on attributes/paths.
orivej has quit [Ping timeout: 264 seconds]
orivej_ has joined #nixos-dev
bennofs_ has joined #nixos-dev
bennofs has quit [Ping timeout: 256 seconds]
globin has quit [Ping timeout: 272 seconds]
globin has joined #nixos-dev
orivej_ has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-dev
cole-h has joined #nixos-dev
<kloenk>
The hash of a nix store path, in what format is it? where does the base32 alphabet comes from? does it have an standart?
<niksnut>
no it's not standard
<kloenk>
niksnut: any reason for this alphabet? just curios
<cole-h>
I've read in the past that some characters were removed to prevent accidental swearing, offensive terms, etc., from showing up in hashes. Don't know how accurate that is, though. Might be covered in the paper?