| 20 Oct 2024 |
 dotlambda | In reply to @artturin:matrix.org
It kinda looks like most of the commits without open-webui in them are package updates to packages which are new to that branch and backported in the same pr Doesn't look like that to me. The worst offender is duckduckgo-search which saw a major version bump and whose dependencies were changed. | 03:31:12 |
 Artturin | In reply to @artturin:matrix.org
It kinda looks like most of the commits without open-webui in them are package updates to packages which are new to that branch and backported in the same pr Only the duckduckgo-search updates might be ineligible but that depends on if there were breaking things in those updates | 03:31:17 |
 titaniumtown (they/them) | In reply to @tomodachi94:matrix.org
Merged :) I immediately made a PR to update it haha: https://github.com/NixOS/nixpkgs/pull/349910
:ppp | 03:31:48 |
 Tomodachi94 | I'll take a look at that too in a few minutes :) | 03:32:16 |
| dotlambda | In reply to @artturin:matrix.org
Only the duckduckgo-search updates might be ineligible but that depends on if there were breaking things in those updates Did you go through the changelogs, especially of the minor rather than patch updates, to check they aren't breaking? I doubt drupol did | 03:32:33 |
| Tomodachi94 | * I'll take a look at that too in a few minutes :) I won't merge it as fast to give the maintainer a chance to review it as well though | 03:32:44 |
| dotlambda | * Did you go through the changelogs, especially of the minor rather than patch updates, to check they aren't breaking? I doubt drupol did.
EDIT: Never mind, there are none that weren't also added in that PR | 03:33:52 |
| dotlambda | In reply to @artturin:matrix.org
Only the duckduckgo-search updates might be ineligible but that depends on if there were breaking things in those updates Not breaking any other packages in the stable channel is not a sufficient condition for a backport. Who knows whether the API changed | 03:35:18 |
 emily | the duckduckgo-search update is valid under "Services that would fail without up-to-date client software, such as spotify, steam, and discord", see e.g. https://github.com/deedy5/duckduckgo_search/pull/216 | 03:39:38 |
| Artturin | In reply to @robert:funklause.de
Not breaking any other packages in the stable channel is not a sufficient condition for a backport. Who knows whether the API changed I now went through duckduckgo-search releases and they seem fine from a glance | 03:40:16 |
| emily | In reply to @emilazy:matrix.org
the duckduckgo-search update is valid under "Services that would fail without up-to-date client software, such as spotify, steam, and discord", see e.g. https://github.com/deedy5/duckduckgo_search/pull/216 ("The old versions stop working, I have to figure out new ways.") | 03:40:51 |
| emily | only other existing packages i see are open-webui itself and shapely | 03:41:04 |
| emily | shapely is a 0.0.1 bug fix release with a small diff, no knowledge/opinion about open-webui | 03:41:23 |
| dotlambda | open-webui is added in that PR, that's fine | 03:41:51 |
| dotlambda | Looks like it's fine after all but such backports really should provide a rationale | 03:42:20 |
| emily | no, that package was preexisting, but everything else at least seems fine | 03:42:25 |
| dotlambda | In reply to @emilazy:matrix.org
no, that package was preexisting, but everything else at least seems fine Oh you're right | 03:43:00 |
| emily | ah, though it was only backported to 24.05 in August | 03:48:16 |
| emily | and the bump was merged the day after | 03:49:00 |
| emily | so not really any break | 03:49:03 |
| emily | https://github.com/NixOS/nixpkgs/pull/332412 was the previous PR with some discussion | 03:49:08 |
| emily | given the august 7 and august 8 merge dates i wouldn't be surprised if the old version never even reached a channel | 03:49:26 |
 SigmaSquadron | crisis averted then? | 03:49:45 |
| emily | i do not personally see any issue currently at least | 03:50:38 |
| Artturin | In reply to @robert:funklause.de
Oh you're right Maybe edit the messages in other channels to be a bit milder since there weren't really any major violations, just some missing motives 😅 | 03:50:39 |
| emily | it's important that we have people check work against our standards but when criticizing people for sloppiness it's especially important to make sure we don't make mistakes from oversights ourselves in the process. could use a little less shoot-first-ask-questions-later | 03:56:28 |
| Tomodachi94 | Does anyone know if stuebinm's nixq has any documentation? Investigating it to see if it can help me semi-automatically do a treewide change (replacing qt5.mkDerivation with stdenv.mkDerivation in most cases). | 04:39:03 |
| Tomodachi94 | Does anyone know if stuebinm's nixq has any documentation? Investigating it to see if it can help me semi-automatically do a treewide change (replacing qt5.mkDerivation with stdenv.mkDerivation in most cases and adding qt5.wrapQt5Apps). | 04:40:51 |
| Tomodachi94 | cc @stue:matrix.org, who might be stuebinm? | 04:45:08 |
| emily | previous discussion fwiw https://matrix.to/#/!kjdutkOsheZdjqYmqp:nixos.org/$BNb-xIiQz6tsVHvZZAAMmVkBV0lGCXig0Ovmoqqf-Rc?via=nixos.org&via=matrix.org&via=nixos.dev | 04:46:46 |
