Synthetica has quit [Quit: Connection closed for inactivity]
ris has quit [Ping timeout: 268 seconds]
ixxie has quit [Ping timeout: 268 seconds]
lovesegfault has quit [Quit: WeeChat 2.7]
<samueldr>
arianvp: pushed an update to cross-system that should build for staging, two new workarounds
orivej has joined #nixos-dev
rajivr___ has joined #nixos-dev
drakonis has quit [Quit: WeeChat 2.6]
bhipple has joined #nixos-dev
bhipple has quit [Remote host closed the connection]
justanotheruser has quit [Quit: WeeChat 2.6]
justanotheruser has joined #nixos-dev
orivej has quit [Ping timeout: 258 seconds]
mjsir911 has quit [Ping timeout: 246 seconds]
mjsir911 has joined #nixos-dev
zarel has quit [Ping timeout: 240 seconds]
Nyanloutre[m] has quit [*.net *.split]
rycee has quit [*.net *.split]
cocreature has quit [*.net *.split]
misuzu has quit [*.net *.split]
genesis has quit [*.net *.split]
thoughtpolice has quit [*.net *.split]
davidtwco has quit [*.net *.split]
arcnmx has quit [*.net *.split]
alienpirate5 has quit [*.net *.split]
yegortimoshenko has quit [*.net *.split]
dywedir[m] has quit [*.net *.split]
Ox4A6F has quit [*.net *.split]
klys has quit [*.net *.split]
Mic92 has quit [*.net *.split]
hexa- has quit [*.net *.split]
fpletz has quit [*.net *.split]
timokau[m] has quit [*.net *.split]
rnhmjoj has quit [*.net *.split]
mkg20001 has quit [*.net *.split]
vaibhavsagar has quit [*.net *.split]
bennofs[m] has quit [*.net *.split]
thefloweringash has quit [*.net *.split]
roberth has quit [*.net *.split]
domenkozar[m] has quit [*.net *.split]
xwvvvvwx has quit [*.net *.split]
simpson has quit [*.net *.split]
flokli has quit [*.net *.split]
capisce has quit [*.net *.split]
Jackneill has joined #nixos-dev
dtz has quit [Ping timeout: 252 seconds]
jonge[m] has quit [Ping timeout: 252 seconds]
layus[m] has quit [Ping timeout: 250 seconds]
jtojnar has quit [Ping timeout: 252 seconds]
worldofpeace has quit [Ping timeout: 252 seconds]
aanderse has quit [Ping timeout: 245 seconds]
ma27[m] has quit [Ping timeout: 246 seconds]
abbradar[m] has quit [Ping timeout: 260 seconds]
Ericson2314 has quit [Ping timeout: 260 seconds]
fpletz has joined #nixos-dev
xwvvvvwx has joined #nixos-dev
hexa- has joined #nixos-dev
capisce has joined #nixos-dev
flokli has joined #nixos-dev
simpson has joined #nixos-dev
tv has quit [Ping timeout: 240 seconds]
hexa- has quit [Max SendQ exceeded]
kraem has joined #nixos-dev
tv has joined #nixos-dev
hexa- has joined #nixos-dev
zarel has joined #nixos-dev
__monty__ has joined #nixos-dev
Mic92 has joined #nixos-dev
klys has joined #nixos-dev
davidtwco has joined #nixos-dev
cocreature has joined #nixos-dev
thoughtpolice has joined #nixos-dev
genesis has joined #nixos-dev
misuzu has joined #nixos-dev
ixxie has joined #nixos-dev
ixxie has quit [Ping timeout: 248 seconds]
ixxie has joined #nixos-dev
ixxie has quit [Ping timeout: 246 seconds]
<arianvp>
Thanks samueldr
<arianvp>
Will try when I get to the office!
ixxie has joined #nixos-dev
ivan has joined #nixos-dev
ixxie has quit [Ping timeout: 245 seconds]
<arianvp>
samueldr: are we sure it's merged yet? I only see something related to pythonPackages.pyparsing
<arianvp>
samueldr: oooh wait to your cross-system repo
<arianvp>
not nixpkgs
<arianvp>
got it
<arianvp>
thanks it compiles!
mkaito has joined #nixos-dev
<mkaito>
infinisil gchristensen about that thing on reddit yesterday, I talked to yorick and we're going to set aside a little time each day to comb through open nixpkgs PRs on company time, whenever possible.
<arianvp>
interestingly I have the opposite experience. Found other distros super chory to contribute to; or even inpenetrable. and nixpkgs has been the only one so far where I actually feel comfortable getting stuff in... :). Not saying your experience isn't true or valid by the way; just found the difference interesting
<arianvp>
(in our experiences; that is)
<Taneb>
Nix has the disadvantage here that it's a) small, and b) weird
<arianvp>
(Arch being one of the exceptions; where packaging is rather easy and streamlined)
<Taneb>
There's not as much knowledge transfer coming to Nix as I imagine there is moving between other package managers, and we don't have the volume to make up for it
<mkaito>
whoops, had the buffer set to notify on highlight
<mkaito>
sorry, didn't mean to ignore anyone
<mkaito>
yes, that's the thing
<arianvp>
But i still don't know after 10 years how to package a deb; for example. after me trying serveral times to learn it
<mkaito>
deb packaging is notoriously obtuse
<mkaito>
when I first came to nixos, I was already a unix graybeard, but nixos changes everything and puts this weird language on top :P it was jarring, and the notoriously abhorrent error reporting didn't help. took a long time for me to get anywhere near proficient with nix and nixos.
<mkaito>
so how does ofborg actually work? For example, I see in this PR[1] that it does a bunch of eval checks, but doesn't actually build anything. Does it run nixos tests? How do I ask it to do things? [1]: https://github.com/NixOS/nixpkgs/pull/78352
<worldofpeace>
Profpatsch: pretty cool, where could addToPassthru be useful in nixpkgs and outside of it? I'm assuming you already have use for this)
<Profpatsch>
worldofpeace: everywhere we add a passthru to an existing derivation. I’ve been using it in various situations.
kenjis1 has joined #nixos-dev
<bennofs>
Profpatsch: I've seen the 'lib.extendDerivation true passthru graalvm8' a few times in nixpkgs, I suppose that's similar?
<bennofs>
btw, what is the difference between overrideAttrs and //?
<bennofs>
and between addToPassthru and //?
__monty__ has quit [Quit: leaving]
<arcnmx>
I believe the difference is that these approaches will preserve what was added when customizing further, where (drv // { something = else; }).overrideAttrs (...) would lose something?
drakonis1 has joined #nixos-dev
kenjis1 has quit [Remote host closed the connection]
<infinisil>
niksnut: rycee: Reminder that the meeting for RFC 42 is in 45 minutes. I'll send a jitsi link before that
kenjis1 has joined #nixos-dev
orivej has joined #nixos-dev
<infinisil>
(link sent in PMs)
<infinisil>
(and via mail)
<rycee>
infinisil: Gotcha.
<Profpatsch>
bennofs: addToPassthru uses both overrideAttrs and //
<Profpatsch>
It’s specialized to the use-case of adding something to passthru.
drakonis1 has quit [Quit: WeeChat 2.6]
avn has joined #nixos-dev
<bennofs>
Profpatsch: I was just wondering what the difference between .overrideAttrs + passthru is compared to just // on the derivation itself
<Profpatsch>
bennofs: oh, good question. I don’t know. But we had overrideDerivation before, which was broken. I think it’s strictness problem.
<Profpatsch>
Once you apply `derivation` to something, it’s done.
<Profpatsch>
In nixpkgs we wrap that to provide magic like passthru.
<Profpatsch>
Which doesn’t change the builder environment, because it’s not passed to derivation in the first place.
<Profpatsch>
And I think overrideAttrs is applied before that, but I’d have to look at the code again to figure out the exact semantics.
<bennofs>
Profpatsch: yeah, but with passthru it's not passed to the builder anyway
<bennofs>
so applying to afterwards should be the same
<arcnmx>
iirc passthru basically just does derivation { ... } // { stuff... } except as mentioned, it composes with customization and preserves the attrs when it's extended in other ways
<bennofs>
but arcnmx is right, if you apply it with // then it will get lost if you .overrideAttrs on the result again, so something like (foo // extraPassthru).overrideAttrs {} will loose the extraPassthru
<bennofs>
interaction of various override mechanisms is broken in many cases :(
<bennofs>
ok, you have convinced me that overrideAttrs is the correct way to do this
<bennofs>
does the nix bot support nix repo search?
ixxie has joined #nixos-dev
WilliButz has quit [Quit: bye]
WilliButz has joined #nixos-dev
ixxie has quit [Ping timeout: 265 seconds]
<Profpatsch>
bennofs: Even if it worked, there might be a question of efficiency. Adding new layers to the fixpoint is usually less efficient than using an existing mechaniscm.
<Profpatsch>
-c
<Profpatsch>
Of course it depends on how big the fixpoint is, but in the worst case, any call to self suddenly has one more indirection
<Profpatsch>
At least that’s my intuition, it might be wrong.
<pie_[bnc]>
cc clever, elvishjerricco, for no particular reason
<elvishjerricco>
Oooh that's good
<pie_[bnc]>
im not figuring out how to put it in nixpkgs though :P and its still gonna bother me that pparently the grub bootloader scrips are happy to fail silently or whatever
<infinisil>
I should make this a functionality of the bot though
<mkaito>
given that no special access is needed, sounds like a good idea. the list of PRs is public anyway.
<infinisil>
Well the script needs a github token to query the API (because I implemented it atop of the graphql api, which doesn't allow unauthenticated requests..)
<mkaito>
infinisil: make a throwaway user and make a token that can read public repos.
<infinisil>
Hmm I think I might implement this now
<mkaito>
we've got a few throwaway users to scope access for various things. sure wish github made this kind of thing easier.
<infinisil>
I guess since I can restrict access for the token pretty well and I could control the bots rate limiting myself too it shouldn't be a problem to use my own token
<mkaito>
I tend to be very careful with putting my own token in places
<mkaito>
I have too much access to sensitive things to even consider risking it :P
<infinisil>
Ah this token would still be private
<Profpatsch>
lol infinisil, why would you open a discussion on THE bikeshed topic there is, which is documentation formats
<infinisil>
Profpatsch: Well *somebody* has to do it!
<infinisil>
And over the last 5 years or so nobody did!
<mkaito>
do you have a minute to talk about our lord and savior, tabs vs spaces? :P
<mkaito>
(in the distance, you hear rabid screeching)
<Profpatsch>
I’m skipping through this thing, and there’s literally no progress being made on anything, except for people giving their opininios
<infinisil>
Profpatsch: Yeah I noticed that too.. It's a bit of a non-productive discussion..
<Profpatsch>
who could have expected that!
<infinisil>
I wish I could edit comments freely, shortening them or marking them as non-relevant/offtopic
<Profpatsch>
If I were to give another opinion, I’d say that this is trying to do instant full change instead of incremental change
<Profpatsch>
In order to achieve anything here, the migration must be gradual.
<infinisil>
Profpatsch: The incremental bit can still be done with the actual implementation of this change
<infinisil>
Nothing says all the docs need to be converted in one go
<infinisil>
It just says what the target of the conversion is
<Profpatsch>
So why not sneakily just select something and start implementing it.
<Profpatsch>
And then ask (non-publicly!!!) if people like it.
orivej has joined #nixos-dev
<infinisil>
That just gives a stuck PR because some core people think it's the wrong move
<Profpatsch>
opinions are for cheap, and people who don’t get anything done have a lot of time writing their opinions.
<infinisil>
Like the org-mode one
<infinisil>
This RFC should give people the direction for where to move without having to worry about it being rejected
<Profpatsch>
Why, then don’t announce it as “replacing the current docbook”, but just do it in a little corner
<Profpatsch>
and then gradually make that corner bigger.
<Profpatsch>
You can see that it works: We have markdown docs!
<infinisil>
Our docs are already split enough
<infinisil>
And with a partial conversion it just means another split
<infinisil>
I really think a single doc format is the way to go
<gchristensen>
+1
<Profpatsch>
infinisil: Do you have a list of all the points that make a solution a no-go?
<gchristensen>
there is a lot of good discussion on the RFC, which I'd recommend you read :)
<infinisil>
Profpatsch: There's a list of requirements in the rfc
<Profpatsch>
gchristensen: haha, lol, I didn’t realize this is #nixos-dev :D
<gchristensen>
where did you think it was? :)
<infinisil>
Hehe
<Profpatsch>
gchristensen: I’d have to kill you … yaddayadda
<gchristensen>
sweet relief!
<Profpatsch>
Two days away from IRC and I forget how to do it.
<pie_[bnc]>
Profpatsch: my excuse is always that i was probably tired :P
<pie_[bnc]>
and have tunnel vision
<Profpatsch>
pie_[bnc]: Those are good links, even if you haven’t looked very deeply into them!
<pie_[bnc]>
Profpatsch: also, the moment when you realize gchristensen is not in [redacted], so why are you reading his messages ;P
<Profpatsch>
ikr
<pie_[bnc]>
you goofed son
<pie_[bnc]>
infinisil: i have no idea what this is, but it says its better... "The fancy mkAliasAndWrapDefsWithPriority thing would be better because it neither generates an error nor does it ignore assignments, but it's very different to use from what people already know, so we'll not do that because of this."
<infinisil>
Ah I meant to link to an example of it
<pie_[bnc]>
so are we ever going to get the better thing?
<infinisil>
pie_[bnc]: Nope
<infinisil>
It's a bit too weird to use
<infinisil>
I updated the comment with a link to an example
<pie_[bnc]>
well ok..
ris has joined #nixos-dev
emily is now known as emilazy
ematrix is now known as emily
Synthetica has quit [Quit: Connection closed for inactivity]
<Profpatsch>
Did we have a way to get a path’s hash from a drv?
<Profpatsch>
huh, phew, not sure I want to depend on that. Our server is still going to send stale content and it breaks if we ever decide to change the reverse proxy server.
<gchristensen>
hum?
<gchristensen>
your proxy isn't e-tag compliant?
<Profpatsch>
gchristensen: No, the static content is served from the web application, and just proxied through nginx.
<Profpatsch>
Maybe we should serve it from nginx instead.
<gchristensen>
ah!
<Profpatsch>
everything’s hard.
<gchristensen>
pretty sure this etag is only set by nginx if serving static files off disk, so this patch shouldn't effect you
<Profpatsch>
yeah, would be very creepy otherwise
<gchristensen>
creepy is a good way to put it
<Profpatsch>
I think I’ll just go for: pass an Option<Etag> to our webserver, and if it’s passed it’s set and if it’s not passed caching is disabled alltogether.
<Profpatsch>
A CAS could do 200s for everything, but that would require rewriting links in our page.
justanotheruser has quit [Ping timeout: 265 seconds]
<mkaito>
oh, that's apparently not supposed to be merged
<mkaito>
still neat that you did that :) I'll be using that tomorrow
<infinisil>
:D
<mkaito>
I'm sure yorick would also find it useful
<infinisil>
I wish this functionality was in github directly..
<infinisil>
"random sort order"
ixxie has joined #nixos-dev
<mkaito>
yeah, I was just thinking about this. if I pick the newest, the oldest will get neglected. but if I pick the oldest, I'll end up trudging through crap from 2015 that isn't even relevant anymore.
evanjs has joined #nixos-dev
tv has joined #nixos-dev
ixxie has quit [Ping timeout: 265 seconds]
ivan has joined #nixos-dev
lopsided98 has joined #nixos-dev
Jackneill has joined #nixos-dev
harrow has joined #nixos-dev
hyperfekt has joined #nixos-dev
orivej has joined #nixos-dev
cptchaos83 has joined #nixos-dev
srk has joined #nixos-dev
xwvvvvwx has joined #nixos-dev
ris has quit [Ping timeout: 258 seconds]
ris has joined #nixos-dev
<clever>
gchristensen: any prometheus exporters for hydra? to replace hydra-send-stats
ixxie has joined #nixos-dev
ixxie has quit [Ping timeout: 240 seconds]
<gchristensen>
yes two, look for "reexporter" in nixos-org-configurations and also look at the prometheuus scrape targets for job-specific exporters
<clever>
ah
kenjis1 has quit [Ping timeout: 260 seconds]
drakonis_ has joined #nixos-dev
ixxie has joined #nixos-dev
ixxie has quit [Ping timeout: 268 seconds]
drakonis_ has quit [Read error: Connection reset by peer]