| 13 Oct 2025 |
 helle (just a stray cat girl) | just shovelling a pile of doc fixes around it seems, lol | 11:54:05 |
|  ekler joined the room. | 12:34:53 |
|  @polytroid:matrix.org left the room. | 13:47:27 |
| ekler | 12:35:04 |
|  azularina joined the room. | 14:01:55 |
| 14 Oct 2025 |
 aloisw | What is going on with the Darwin machine with IP address 2001:bc8:38ee:100:6666::2? I still cannot reach it (SSH times out and it also doesn't react to pings). | 08:22:28 |
 delroth | aloisw: is that the same as m1@build01.aarch64-darwin.lix.systems? if so it's currently down for unknown reasons and needs a manual reboot, discussed here: https://zulip.lix.systems/#narrow/channel/7-T-infra/topic/SCW.20macOS.20seems.20down | 10:08:35 |
 raitobezarius | no, 6666 is one of the AFNix macOS builder | 10:08:55 |
| delroth | (the DNS zone only has the v4 for that host so I don't know if we're talking about the same one) | 10:09:01 |
| raitobezarius | we didn't setup yet access to these mac for interactive remote building | 10:09:14 |
| delroth | ahh, thanks | 10:09:36 |
| raitobezarius | nw | 10:09:39 |
| delroth | I got confused since the whois resolves to scaleway | 10:09:42 |
| raitobezarius | 2001:bc8:38ee::/48 is par01 | 10:10:01 |
| delroth | but wait | 10:10:49 |
| delroth | the afnix macos builders don't have v6? | 10:10:53 |
| raitobezarius | not yet | 10:10:59 |
| raitobezarius | the initial setup had v6 | 10:11:02 |
| delroth | where's that IP from then? | 10:11:11 |
| raitobezarius | then we switched them over v4 because we needed to complete the MDM process | 10:11:12 |
| raitobezarius | it's from the initial setup before MDM | 10:11:19 |
| raitobezarius | i can reintroduce the IPv6 once I have some time | 10:11:33 |
| delroth | I'm just confused how aloisw ended up with it then :p | 10:11:37 |
| delroth | ok | 10:11:37 |
| delroth | I guess I don't need the full history lesson to accept current facts | 10:11:47 |
| raitobezarius | before MDM, I gave aloisw access to one of the box so they could do some testing | 10:11:52 |
| delroth | makes sense, thanks | 10:12:01 |
| raitobezarius | we can move this over AFNix discussion | 10:12:10 |
| raitobezarius | (https://zulip.afnix.fr/#narrow/channel/4-infra/topic/macOS.20remote.20building.20access for those who wants to follow) | 10:13:03 |
 emily | has there been any more thought given re: builtins.convertHash / https://git.lix.systems/lix-project/lix/issues/602? I know making it a more general radix conversion function was discussed, but IMO the Base32 and Base64/SRI hash formats are quirky enough that they really don't fit elegantly into a radix conversion model, especially since they are too large to round-trip via Nix integers, which would otherwise be the natural way to express radix conversion functions.
not saying it has to have the exact same interface but I am not convinced that the full general base conversion mechanism is necessary/desirable for it, and the -2 review on https://gerrit.lix.systems/c/lix/+/2585 shows that it would require a lot of contortions to make it useful for the actual hash conversion use case, which seems to have been useful both in third-party repos and has caused a hash mismatch in Nixpkgs even with an attempt to do feature detection https://github.com/NixOS/nixpkgs/pull/451929.
| 15:38:52 |
