| 10 Jan 2025 |
 Niklas Korz | https://wiki.nixos.org/wiki/MatrixRooms also does not list one, so I guess not 😄 | 10:32:14 |
| Niklas Korz | So I'll just ask here (also it's probably more of a stdenv question than a Ruby-specific question):
pkgs/development/ruby-modules/bundled-common/default.nix has special handling for gemset to be of type set instead of string / path, which I'd like to take advantage of to make it easier patching the gemset attribute set from overrideAttrs and passing it to bundlerEnv. This works fine.
Now to allow the override I tried setting gemset as an attribute on the stdenv args and then reference it as finalAttrs.gemset in bundlerEnv. Thid doesn't work as stdenv errors about gemset not being a string. I know I could set gemset as a package parameter instead and use override to override gemset as a whole, but the idea of using finalAttrs and gemset as an attribute set was that I could access prevAttrs.gemset in overrideAttrs. Is there a way to pass along non-string attributes in finalAttrs?
| 10:39:34 |
| Niklas Korz | * So I'll just ask here (also it's probably more of a stdenv question than a Ruby-specific question):
pkgs/development/ruby-modules/bundled-common/default.nix has special handling for gemset to be of type set instead of string / path, which I'd like to take advantage of to make it easier patching the gemset attribute set from overrideAttrs and passing it to bundlerEnv. Setting gemset to an attr set works fine, so this is not the problem.
Now to allow the override I tried setting gemset as an attribute on the stdenv args and then reference it as finalAttrs.gemset in bundlerEnv. Thid doesn't work as stdenv errors about gemset not being a string. I know I could set gemset as a package parameter instead and use override to override gemset as a whole, but the idea of using finalAttrs and gemset as an attribute set was that I could access prevAttrs.gemset in overrideAttrs. Is there a way to pass along non-string attributes in finalAttrs?
| 10:40:16 |
 Gaétan Lepage | Looks like GHA are struggling right now... | 10:50:51 |
| Niklas Korz | ok for my usecase it's not relevant anymore, but would be interesting to know nonetheless 😄 | 11:06:59 |
 @nhp:matrix.org | how would I go about submitting an update to a package (ares) that needs a new dependency (librashader)? I submitted a PR to package librashader, but updating ares depends on that being merged, so should I just wait or is there a way to submit a PR that depends on another PR? | 16:18:23 |
 K900 | You can just put both in one PR | 16:18:56 |
| @nhp:matrix.org | ah, okay - should I just close my existing one or is it better to edit it? | 16:30:18 |
 emily | up to you | 16:37:39 |
| emily | you can do stacked PRs in GitHub only if you can push branches directly to the repo :( | 16:37:55 |
| emily | (by directing one PR at another PR's branch; when the other PR is merged the target branch gets updated) | 16:38:06 |
 Randy Eckenrode | Isn’t there some weirdness with that if you aren’t targeting master with the first PR? We ran into that with the Darwin SDK stuff IIRC. | 17:19:44 |
| emily | I think that was because the PR was from your fork | 17:22:17 |
| emily | we just had a mirror branch set up but it turned out to not be enough | 17:22:23 |
| emily | but maybe staging could matter too? | 17:22:27 |
|  @sigmasquadron:matrix.org changed their display name from SigmaSquadron (Away until 2024-01-12) to SigmaSquadron (Away until 2024-01-11). | 22:15:41 |
| @sigmasquadron:matrix.org changed their display name from SigmaSquadron (Away until 2024-01-11) to SigmaSquadron (Away until 2025-01-11). | 22:43:10 |
| 11 Jan 2025 |
|  @orzklv:matrix.org left the room. | 04:07:51 |
|  Sinan (comfy) changed their profile picture. | 11:36:33 |
| @sigmasquadron:matrix.org changed their display name from SigmaSquadron (Away until 2025-01-11) to SigmaSquadron. | 15:06:32 |
 @marie:marie.cologne | i need some help with jj, I want to rebase some work to staging, I thought I could just do jj rebase -d staging, but that gives me Error: Commit 302c5392f503 is immutable. I tried a bunch of things, but I didn't figure out the correct thing. What do I need to do? | 16:11:26 |
| @marie:marie.cologne | 302c5392f503 is my local master bookmark where my work is on top | 16:12:44 |
 mjm | i'm guessing that master is not merged into staging, so it's trying to include that in the rebase | 16:13:03 |
| emily | perhaps pass -s <base commit of your work>. by default it will infer the stack to rebase from your current branch of work | 16:13:18 |
| mjm | you probably need -s to be more specific about which change to rebase | 16:13:16 |
| emily | which if you haven't pushed may not be accurate | 16:13:26 |
| emily | I would also suggest -d 'fork_point(master, staging)' when possible, so that you can test on a revision close to master that still merges into staging | 16:13:44 |
| emily | (this is the equivalent of the git merge-base suggestion in CONTRIBUTING.md) | 16:13:53 |
| mjm | woah there's a fork_point function in there? | 16:14:08 |
| mjm | i had written my own merge_base revset alias for that lol | 16:16:38 |
