| 22 Jan 2025 |
 waltmck | build just failed same as before and memory use never exceeded 40G | 03:41:15 |
 atemu12 | Okay took a look at the error and it's quite clear; wouldn't you say? ccache: error: Failed to create directory /var/cache/ccache/tmp: Not a directory | 04:01:25 |
| atemu12 | You turned on ccache and didn't configure it properly | 04:01:44 |
| atemu12 | You need to pierce the sandbox for ccache to make any sense | 04:02:42 |
| atemu12 | Documentation on that could be better but it lays out the requirements | 04:02:58 |
| waltmck | ah, that was very silly | 04:21:21 |
| waltmck | thank you! | 04:21:23 |
| 26 Jan 2025 |
|  cadecadet joined the room. | 23:11:45 |
| 27 Jan 2025 |
|  @panxer:matrix.org left the room. | 01:57:16 |
|  autra joined the room. | 09:17:52 |
| autra | Hi! I'm trying to setup robotnix to build latests los 22 for my fp3. I've added "FP3" to flavors/lineageos/supported_devices.toml and I'm executing flavors/lineageos/update.sh lineage-22.1. Some things seem to work correctly (flavors/lineageos/device-metadata.json and flavors/lineageos/lineage-22.1/repo.json are updated), but this script keeps failing with errors like:
subprocess.CalledProcessError: Command '['git', 'ls-remote', 'https://android.googlesource.com/platform/external/rust/crates/tower']' returned non-zero exit status 128.
(never the same repo).
What's going on? I feel like I'm getting rate limited (especially because there are 429 errors before)... Is it the case?
| 09:23:00 |
| atemu12 | Skip the repo update | 09:23:38 |
| atemu12 | New devices only need the device dir updates | 09:23:54 |
| atemu12 | But yes, that's likely just a rate limit | 09:25:15 |
| autra | There is way to authenticate to android.googlesource.com to avoid that, but I don't know if robotnix supports that or not. | 09:27:28 |
| atemu12 | You don't need to | 09:27:40 |
| autra | So I don't use update.sh? (the command that fails is /mk_repo_file.py in this script) | 09:28:12 |
| atemu12 | You don't need to contact any googlesource.com sources | 09:28:16 |
| atemu12 | Uncomment the repo update | 09:28:26 |
| atemu12 | If you're motivated, you could build in an option to do that; PRs welcome | 09:29:23 |
| autra | just the "repo init" in mk_repo_file.py? | 09:32:28 |
| autra | apparently not | 09:32:52 |
| atemu12 | No the entire step | 09:32:55 |
| autra | I've commented the call to mk_repo_file.py entirely | 09:33:05 |
| atemu12 | Yup | 09:33:11 |
| autra | The rest of the tree is somewhere else? In cachix? | 09:33:35 |
| atemu12 | The rest of what tree? | 09:33:52 |
| atemu12 | The cachix contains nothing of interest FWIW | 09:36:44 |
| autra | ok I'm just suprised by what you said (new devices only need the device dir update), because my experience with android build is that you need the whole lineage os tree to build, but I guess I'll see all these downloads later ;-) | 09:53:01 |
| atemu12 | The updater really is just thatz an updater | 09:53:53 |
