1 Dec 2022 |
| hexa changed their display name from hexa to hexa (22.11 now). | 13:09:19 |
| hexa changed their display name from hexa (22.11 now) to hexa. | 14:39:37 |
3 Dec 2022 |
| K900 ⚡️ joined the room. | 16:37:39 |
K900 ⚡️ | qbittorrent is also not getting updated for some reason | 16:38:16 |
| @trofi:matrix.org joined the room. | 16:38:20 |
hexa | https://r.ryantm.com/log/re2/ | 16:38:22 |
hexa | looks like re2 has logs until 2022-11-02 | 16:38:28 |
hexa | oh you meant the global log stopped in july | 16:39:58 |
@trofi:matrix.org | It's both: re2 was released a few days ago, but attempts seems to have stopped a month ago. And global log (where I hoped to see any attempts, or backlog, or something) dud not update since July. | 16:42:16 |
5 Dec 2022 |
| raphi joined the room. | 18:05:30 |
| ribosomerocker changed their display name from mon_aaraj to ribosomerocker. | 22:45:28 |
6 Dec 2022 |
raphi | if a derivation is split across multiple files (fetchGitHub in one, version/hash in the other), will r-ryantm attempt to auto-update it?
Example: generic.nix and 2.nix here: https://github.com/NixOS/nixpkgs/pull/204658/commits/9cbf5af89e54f3b6ecd271f6d976aeb69d8bba53 - in this case, i'm wondering if it's still neccessary to disable r-ryantm using the "no auto update" comment. | 08:22:49 |
Sandro 🐧 | the comment about dev releases is still true? | 10:21:55 |
Sandro 🐧 | not sure if the auto updater can handle this. In question I would just see what happens. | 10:22:21 |
ryantm | raphi: I'd expect it to try and fail. If you want it to work, you should probably add a passthru.updateScript. | 11:51:24 |
13 Dec 2022 |
hexa | Redacted or Malformed Event | 01:46:04 |
hexa | maybe a test not passing should be more prominent, see https://github.com/NixOS/nixpkgs/pull/189801#issuecomment-1347629531 | 01:46:46 |
ryantm | Now that we do less checks, we could probably stop hiding them behind details. | 02:05:47 |
15 Dec 2022 |
| ribosomerocker changed their profile picture. | 19:53:02 |
18 Dec 2022 |
@trofi:matrix.org | https://r.ryantm.com/log/strace/ <- did not run since January? Noticed there was a release 8 days ago, filed update PR manually. But I'm wondering if it's expected. | 12:56:52 |
ryantm | trofi: Hmm, yeah that's bad. There's obviously problems with the bot, since Repology had newer versions between then and now. https://repology.org/project/strace/history | 22:28:20 |
23 Dec 2022 |
@trofi:matrix.org | Trying to run nixpkgs-update locally to see how it would handle gawk . My attempt:
$ ./result/bin/nixpkgs-update update-batch --attrpath gawk
Configured Nixpkgs-Update Options:
----------------------------------
GitHub User: trofi
Send pull request on success: NO
Batch update: YES
Calculate Outpaths: NO
CVE Security Report: NO
Run nixpkgs-review: NO
Nixpkgs Dir: "/home/slyfox/.cache/nixpkgs"
update info uses attrpath: YES
----------------------------------
nixpkgs-update: Prelude.head: empty list
Is there a verbose mode maybe to figure out where that Prelude.head: empty list come from?
| 19:18:33 |
ryantm | It's probably failing to find the GitHub token @trofi | 19:42:37 |
@trofi:matrix.org | Aha | 19:43:15 |
@trofi:matrix.org | Looks like my token is read fine. Instead it fails at https://github.com/ryantm/nixpkgs-update/blob/main/src/Update.hs#L768 . Looks like parseUpdates updateInfo only contains [Left "Unable to parse update: gawk"] for such command line. | 21:33:27 |
@trofi:matrix.org | Looking at the option parsing update-batch --attrpath gawk is still unacceptable and requires versions passed: https://github.com/ryantm/nixpkgs-update/blob/main/src/Utils.hs#L233. I naively expected versions to be extracted from nixpkgs checkout and packages-to-update.txt . | 21:44:40 |
24 Dec 2022 |
| @omlet:matrix.org left the room. | 00:22:26 |
30 Dec 2022 |
| Federico Damián Schonborn joined the room. | 00:21:15 |
9 Feb 2023 |
| @rdfg77:kde.org left the room. | 06:47:10 |
14 Mar 2023 |
| pbsds joined the room. | 15:31:26 |