| 25 Aug 2025 |
 raitobezarius | for now, every little energy i have is spent on durable automation | 17:35:04 |
| raitobezarius | (because it makes me happier) | 17:35:11 |
| 26 Aug 2025 |
 jade_ | I have given a bunch of people including emily, thubrecht , EsperLily [she/her] , just1602, Linux Hackerman Code-Review+2 access as it seemed silly that y'all didn't have it. in practice this just lets you stamp packaging or docs due to our CODEOWNERS configuration, but if you want to take responsibility for reviews on more of the tree feel free to send a CL to add yourself to CODEOWNERS in your areas.
the people in this list were selected basically arbitrarily based on a subset of people who I have noticed has a very consistent pattern of having good taste and who have some commits since May. most have commit access in our other repos as well already. if anyone else doesn't have and wants +2 access please let me know and i can raise it with the core team.
| 01:02:40 |
 EsperLily [she/her] | this reminds me that i should be more proactive about looking for stuff to review | 01:17:11 |
 just1602 | Oh thanks for the trust !
Also, I started a new job today, and tomorrow I plan to start my day by switching all tne nixos servers I'm now responsible to lix 🎉 | 02:27:52 |
| raitobezarius | @qyriad:katesiria.org: I think I made an obvious mistake in the overlay defn | 12:31:22 |
 Qyriad | I think likely | 12:39:45 |
| raitobezarius | Can you send a PR to replace final by prev in the docs? | 12:56:21 |
| Qyriad | oh wait is it just in the docs? | 16:45:17 |
| raitobezarius | Yep | 17:08:19 |
| 27 Aug 2025 |
|  Tran Minh Hieu joined the room. | 02:45:11 |
|  crop joined the room. | 21:00:47 |
| 28 Aug 2025 |
| just1602 | I gave it a shot and bump nixpkgs to 25.05 for the lix 2.92 branch: https://gerrit.lix.systems/c/lix/+/4120 | 04:02:45 |
| just1602 | I hope it's gonna help until we get renovate to do all the heavy lifting for us :D | 04:03:33 |
 Sergei Zimmerman (xokdvium) | So on 2.93.3 when I do nix flake metadata path:%00 it starts gobbling up all RAM. | 10:06:35 |
| Sergei Zimmerman (xokdvium) | For context, I was doing URL cleanups in CppNix and added validation code for this sort of thing. That seems to be fucked everywhere. Percent encoded NUL bytes are a menace | 10:08:24 |
 helle (just a stray cat girl) | nul bytes in general are a menace, ouch | 10:09:16 |
| Sergei Zimmerman (xokdvium) | I wonder what lix is doing here to gobble up gigabytes of ram in seconds | 10:09:39 |
| Sergei Zimmerman (xokdvium) | Nix just crashes with a stack overflow | 10:09:54 |
| helle (just a stray cat girl) | I have my suspicions | 10:10:01 |
| Sergei Zimmerman (xokdvium) | Coroutine stack frames it seems | 10:11:12 |
| helle (just a stray cat girl) | either stack or indefinitely growing the output array due to a lack of correct termination of the decode | 10:11:43 |
| Sergei Zimmerman (xokdvium) | Hm yeah seems to be growing a std::string in list called from dumpSingle | 10:19:00 |
 piegames | In reply to @xokdvium:matrix.org
For context, I was doing URL cleanups in CppNix and added validation code for this sort of thing. That seems to be fucked everywhere. Percent encoded NUL bytes are a menace What the hell even is this percent encoding 👀 | 13:01:27 |
| piegames | Oh wait that's an URL encoding | 13:31:01 |
| Sergei Zimmerman (xokdvium) | In reply to @piegames:flausch.social
Oh wait that's an URL encoding tomayto, tomahto | 13:34:30 |
| piegames | Dunno, it's just that I forgot Nix uses URLs for Flakes here, so I was bracing for yet another ad-hoc escaping mechanism (the ones within the language are not well designed) | 13:39:02 |
 aloisw | There are also other places where shitty URL parsers that don't handle percent-encoding properly are used, for example the remote builders. | 15:22:44 |
| Sergei Zimmerman (xokdvium) | In lix maybe, don't imagine this code was touched much since 2.18 | 16:10:24 |
| aloisw | Are you saying this is fixed in nix? | 16:48:32 |
