| 6 Oct 2021 |
| Rosario Pulella changed their display name from Rosuavio to Rosario Pulella. | 10:44:57 |
| kamadorueda joined the room. | 22:33:53 |
| 11 Oct 2021 |
cornu | My latest build fails with "The option `source.dirs.source.dirs' does not exist". | 11:21:44 |
cornu | It seems "source.dirs" is doubled in https://github.com/danielfullmer/robotnix/blob/96c0eee36ba725b37a565eea0de21f50f9fae71a/modules/apps/updater.nix#L66 | 11:22:04 |
danielrf | cornu: Thanks for reporting. Just fixed in the latest commit | 16:24:05 |
hmenke | There has also been a release of GrapheneOS. https://grapheneos.org/releases#2021100606 | 17:35:31 |
danielrf | hmenke: Yeah this release it a bit of a weird one since the kernel is built using repos from the android 12 manifest tags, while the rest of the source is built using the android 11 manifest tags | 17:36:55 |
danielrf | I originally thought it was just going to stay in the beta channel upstream so I could just skip this one, but since they've pushed it on their stable channel too I should probably try to build this as well | 17:37:59 |
danielrf | I should have some time later tonight to work on it | 17:38:05 |
danielrf | * hmenke: Yeah this release it a bit of a weird one since the kernel is built using repos (e.g. clang) from the android 12 manifest tags, while the rest of the source is built using the android 11 manifest tags | 17:38:26 |
danielrf | OK, pushed new tag grapheneos-2021100606. Tested working on crosshatch and sunfish | 23:06:25 |
| 12 Oct 2021 |
| kamadorueda left the room. | 01:52:47 |
jack | In reply to @danielrf:matrix.org OK, pushed new tag grapheneos-2021100606. Tested working on crosshatch and sunfish Tested working on redfin! | 15:51:46 |
Yuka (she/her) | In reply to @danielrf:matrix.org OK, pushed new tag grapheneos-2021100606. Tested working on crosshatch and sunfish Tested working on bramble | 15:57:02 |
| 13 Oct 2021 |
hmenke | Does anyone have an estimate of how much time is actually saved by using ccache? I'd really like to run my robotnix builds unprivileged but currently my use of ccache is preventing that. | 15:08:48 |
danielrf | I don't have any concrete number off hand, but my impression was that it mattered a lot more when I was running a low-core count machine compared to my high-core count machine--both in absolute and percentage terms | 19:53:11 |
danielrf | if I had to guess, I'd say 10-20% time savings on my high core count machine, 50% time savings on my low core count machine | 19:53:50 |
danielrf | I think a good part of this difference is just because on high core counts, the build more constrained by I/O, speed of copying files, etc.--also the initial build bootstrapping (before ninja starts) isn't massively parallelized. | 19:57:05 |
| 15 Oct 2021 |
| Xe (xe/they) changed their profile picture. | 19:14:37 |
| 18 Oct 2021 |
kranzes | Hey im getting this error when trying to build with microg.enable | 21:01:12 |
kranzes | FakeStore-v0.1.0.apk> curl: (22) The requested URL returned error: 403 | 21:01:14 |
kranzes | error: builder for '/nix/store/lhl2ny995l5pgangfsl4mvlz5kdnzik1-FakeStore-v0.1.0.apk.drv' failed with exit code 1 | 21:01:26 |
kranzes | anyone else experiencing this? | 21:01:32 |
kranzes | i rolled back my nixos system to an older nixos-unstable version | 21:11:56 |
kranzes | and it worked | 21:11:59 |
kranzes | i am not 100% sure if that was the issue but sure | 21:12:15 |
kranzes | i guess... | 21:12:17 |
| 19 Oct 2021 |
hmenke | Has anyone ever attempted to build robotnix images on Darwin? I'm curious whether it works at all. | 17:50:48 |
danielrf | I doubt it would work with robotnix as-is, since I've assumed we're building x86_64 everywhere. | 17:52:42 |
danielrf | But since the aosp build system does look to support Darwin, it might be possible to add support in robotnix too | 17:53:07 |