Sender | Message | Time |
---|---|---|
15 Jan 2025 | ||
22:06:47 | ||
22:07:30 | ||
16 Jan 2025 | ||
hey, is any action needed on nixpkgs-update's part when repology changes the name of a package? pebble had a name clash between two separate projects, which was causing build failures. now they've renamed the two projects to different names, and pebble doesn't exist anymore: https://repology.org/project/pebble/versions | 09:28:53 | |
* hey, is any action needed on nixpkgs-update's part when repology changes the name of a package? pebble had a name clash on repology between two separate upstream projects, which was causing build failures. now they've renamed the two projects to different names, and pebble doesn't exist anymore: https://repology.org/project/pebble/versions | 09:29:11 | |
so basically, we should be tracking this one: https://repology.org/project/pebble-acme-test-server/versions | 09:30:01 | |
09:38:13 | ||
19 Jan 2025 | ||
Setting null for updateScript doesn't work (and can sometimes break the bot). This should work for the updateScript:
Would also need to add rquickshare-legacy to https://github.com/nix-community/nixpkgs-update/blob/main/src/Skiplist.hs, currently it isn't otherwise possible to tell the bot to ignore the legacy version without splitting them into separate packages. | 01:34:52 | |
I didn't notice that there were links in the topic. I keep https://nix-community.org/update-bot/ updated with links for the bot but I'll need to get someone to update the topic or give me permissions. | 01:43:13 | |
No, shouldn't need to do anything as the package versions will be correct after the rename. | 01:49:23 | |
You'll need to use version-regex in the nix-update-script. | 04:14:51 | |
I don't see much to be gained from this feature, most of the time it will only save a day or so. | 04:19:23 | |
23 Jan 2025 | ||
08:09:31 | ||
24 Jan 2025 | ||
15:52:04 | ||
Hey, I merged https://github.com/NixOS/nixpkgs/pull/376046 yesterday, which added updateScript to azure-cli-extensions, but it looks like it's not working in nixpkgs-update.
but the log says there was no diff: https://nixpkgs-update-logs.nix-community.org/azure-cli-extensions.ssh/2025-01-24.log | 16:01:49 | |
The bot checks for new update scripts twice a day and updates are run on the merge base of master and staging so sometimes the the bot attempts an update script that doesn't work on the current merge base. | 23:39:57 | |
25 Jan 2025 | ||
Makes sense, thanks. Will wait and observe. :) | 09:00:11 | |
27 Jan 2025 | ||
07:39:29 | ||
Can someone help me understand by the bot created https://github.com/NixOS/nixpkgs/pull/374876 ? (downgrade filebeat 8 -> filebeat 7) -- is it because I, in a misguided attempt to stay consistent with the existing I tried loading up nixpkgs-update in a ghci repl, but it appears to indicate I'm good on that front,
How do I fix this? :) | 08:21:04 | |
* Can someone help me understand why the bot created https://github.com/NixOS/nixpkgs/pull/374876 ? (downgrade filebeat 8 -> filebeat 7) -- is it because I, in a misguided attempt to stay consistent with the existing I tried loading up nixpkgs-update in a ghci repl, but it appears to indicate I'm good on that front,
How do I fix this? :) | 08:24:43 | |
In reply to @srhb:matrix.orgProbably you need to fix the updateScript, that is what the bot used. | 08:54:21 | |
Ah, yes. I'll have a look at that. | 08:54:43 | |
Yep, thanks, think I fixed it, https://github.com/NixOS/nixpkgs/pull/377256 | 09:13:52 | |
28 Jan 2025 | ||
could we reopen https://github.com/Mic92/nixpkgs-review/issues/430 please? see https://github.com/NixOS/nixpkgs/pull/377508#issuecomment-2618905015 | 20:34:31 | |
2 Feb 2025 | ||
16:04:48 | ||
3 Feb 2025 | ||
16:25:35 | ||
5 Feb 2025 | ||
01:49:47 | ||
6 Feb 2025 | ||
What's the usual practice for nested attributes where I'd like to expose an update script? For instance, we want to expose a single attribute in androidenv , which is an attrset, not a derivation itself. | 03:29:53 | |
Provided, this is also a custom update script not using nix-update or something else. | 03:30:42 | |
Do I need to expose the passthru on the top level attribute? Is the solution a well placed recurseIntoAttrs? Never done this before, so unsure. :-) | 03:34:15 | |
is this what are you looking for? | 07:43:46 |