!VRULIdgoKmKPzJZzjj:nixos.org

Nix Package Manager development

837 Members
For people hacking on Nix: https://github.com/NixOS/nix Nix maintainers can be reached here.180 Servers

Load older messages


SenderMessageTime
27 Jul 2025
@chipset:boykisser.infochipsetaleksana 🏳️‍⚧️ (force me to bed after 18:00 UTC): 11:50:47
@aleksana:mozilla.orgaleksana 🏳️‍⚧️ (force me to bed after 18:00 UTC)
In reply to @chipset:boykisser.info
aleksana 🏳️‍⚧️ (force me to bed after 18:00 UTC):
What?
12:45:29
28 Jul 2025
@mightyiam:matrix.orgShahar "Dawn" OrIs there no meeting now?12:25:03
@xokdvium:matrix.orgSergei Zimmerman (xokdvium)
In reply to @mightyiam:matrix.org
Is there no meeting now?
The hacking sesh got moved, though the calendar didn’t get updated. There’s a crab fit in the meeting pad.
12:44:45
@Ericson2314:matrix.orgJohn Ericson https://github.com/NixOS/nix/pull/13263 Sergei Zimmerman (xokdvium) do you have any ideas on how we should benchmark this? 14:14:18
@Ericson2314:matrix.orgJohn Ericson that was Eelco's worry before 14:14:28
@Ericson2314:matrix.orgJohn Ericson I am not opposed to std::variant<json, string> as a kind of "parse on demand" data structure 14:14:54
@Ericson2314:matrix.orgJohn Ericsonbut it would be good to know whether there is actual a perf issue or not14:15:13
@Ericson2314:matrix.orgJohn EricsonI would think that when .drv is read, it is usually also built, so the parsing will happen anyways, eventually14:15:37
@Ericson2314:matrix.orgJohn Ericsonlikewise when drv is created in the evaluator, we'll have a json object, and end to serializing it, either way14:15:58
@xokdvium:matrix.orgSergei Zimmerman (xokdvium)Microbenchmarks with google/benchmark?14:16:08
@Ericson2314:matrix.orgJohn Ericsonwell my argument above is microbenchmarks would kinda miss the point14:16:42
@Ericson2314:matrix.orgJohn Ericsonyes, parsing does become more expensive14:16:48
@Ericson2314:matrix.orgJohn Ericsonbut other things become less expensive14:16:53
@Ericson2314:matrix.orgJohn Ericsonand it should be some O(drvs) thing either way14:17:06
@xokdvium:matrix.orgSergei Zimmerman (xokdvium)
In reply to @Ericson2314:matrix.org
well my argument above is microbenchmarks would kinda miss the point
But that change would affect store derivation instantiations, right? Or is that not the case?
15:54:37
@Ericson2314:matrix.orgJohn Ericsonwhat do you mean by instantiations?16:05:59
@Ericson2314:matrix.orgJohn Ericsoncreating them in the evaluator?16:06:05
29 Jul 2025
@joerg:thalheim.ioMic92 John Ericson: Sergei Zimmerman (xokdvium) https://github.com/NixOS/nix/pull/13569 07:42:09
@joerg:thalheim.ioMic92Microbenchmarks for derivation parsing07:42:58
@Ericson2314:matrix.orgJohn Ericson thanks Mic92! 14:58:04
31 Jul 2025
@sammy:cherrykitten.devSammy (It/Its) left the room.10:11:40
29 Jul 2025
@Ericson2314:matrix.orgJohn Ericsonhttps://github.com/NixOS/nix/issues/13570 issue I am curious about14:58:08
@joerg:thalheim.ioMic92Left a comment.15:17:42
@Ericson2314:matrix.orgJohn Ericsonthanks!15:24:40
@Ericson2314:matrix.orgJohn Ericsondidn't think of that angle15:24:43
@Ericson2314:matrix.orgJohn Ericson Mic92: any idea where we would get a nice meaty real-world structured attrs derivation? 20:17:00
@Ericson2314:matrix.orgJohn EricsonI'll need that to test my PR20:17:26
@Ericson2314:matrix.orgJohn Ericsonnevermind20:35:30
@Ericson2314:matrix.orgJohn Ericson firefox.overrideAttrs (_: { __structuredAttrs = true; }) that works 20:35:44

Show newer messages


Back to Room ListRoom Version: 6