| 4 Jul 2025 |
 Zhaofeng Li | There is internet sharing, did it support IPv6 before? | 18:43:11 |
 emily | I've been using my MacBook Pro as an AP in lieu of having the BPI-R4 fully set up yet... | 18:43:17 |
| emily | but I guess it's probably still the R4 doing the RAs there? | 18:43:51 |
| Zhaofeng Li | A few days late: not the bootloader-unlockable iPhone people are looking for, but it's a Macbook with an iPhone chip
https://www.macrumors.com/2025/06/30/new-macbook-with-a18-chip-spotted/
| 18:46:32 |
 Randy Eckenrode | In reply to @zhaofeng:zhaofeng.li
There is internet sharing, did it support IPv6 before? It could be doing its own thing. rtadvd kind of sucks, especially if hasn’t been synced with FreeBSD recently. | 18:48:33 |
| Randy Eckenrode | In reply to @emilazy:matrix.org
but I guess it's probably still the R4 doing the RAs there? Probably your router is unless you have your MBP set up as a router (assuming IPv6 at all). | 18:49:11 |
| Randy Eckenrode | In reply to @reckenrode:matrix.org
It could be doing its own thing. rtadvd kind of sucks, especially if hasn’t been synced with FreeBSD recently. IME systemd-networkd > radvd > rtadvd. | 18:49:58 |
| 5 Jul 2025 |
| Randy Eckenrode | I’m not rebasing on staging this week to focus on splitting up my branch. It’s time to get my wrapper changes up for review, bump the minimum to 14.0, and update the SDK and source releases to 15.5. | 02:16:07 |
| Randy Eckenrode | In that order. I made no effort to make the 15.5 source releases build on anything older than 14.0. In fact, I deleted the workarounds to enable that. | 02:16:42 |
| Randy Eckenrode | https://github.com/NixOS/nixpkgs/blob/2ba055ea870e8552ac01ffd2e476db05b104a026/pkgs/by-name/li/libblocksruntime/package.nix | 02:35:20 |
| Randy Eckenrode | How many different places is this stuff packaged? I think there’s another libdispatch package as well. | 02:35:38 |
| Randy Eckenrode | No, it’s just there’s a top-level swift-corelibs-libdispatch and swiftPackages.Dispatch. | 02:38:17 |
| Randy Eckenrode | The former is an alias for the latter. | 02:38:30 |
 tiferrei | Hello! I have nix (and nix-darwin, etc.) on my mac tracking nixpkgs-unstable. For some more stability, I'd like to move to one of the stable branches. I am wondering if 1) this is generally safe to do, and 2) is nixpkgs-25.05-darwin usually as "backported to" as nixos-25.05? Thanks! | 15:08:48 |
 K900 | "Stability" as in less bugs? | 15:09:11 |
| K900 | Because that's not really what those are for | 15:09:17 |
| tiferrei | I get quite frequent breaking changes on packages that seem to be updated quite quickly in nixpkgs-unstable | 15:09:45 |
| tiferrei | After a day or so someone makes a PR fixing it, but id like to avoid these surprises | 15:10:06 |
| K900 | The solution for that is usually to just not update when something fails to build | 15:11:22 |
| emily | the same things are backported | 15:12:27 |
| emily | both of them are fed by Hydra builds of the same release branch | 15:12:41 |
| emily | it is generally safe to move between branches | 15:12:54 |
| tiferrei | Got it, thanks! | 15:13:45 |
| emily | there is also stuff like nix-weather if you want to stay on unstable but get an idea of breakage | 15:14:26 |
| emily | but yes it's also perfectly valid to just not bump if there are build errors | 15:14:34 |
| emily | ymmv | 15:14:35 |
| tiferrei | I see. I currently have a CI bot that just updates my lock every so often. Whenever i remember to pull i bring these changes in. perhaps this is just not advised to update that often | 15:16:04 |
| emily | it's totally fine to update often | 15:16:45 |
| K900 | Maybe build your config in CI and don't push when it doesn't build? | 15:16:49 |
| emily | maybe you could build your system configs in CI and then it'll be obv– yeah | 15:16:53 |
