| 18 Nov 2025 |
 raitobezarius (DECT: 7248) | but ideally, a solution would take into account this branching model or mesh with the releng process | 17:00:48 |
 leona | e.g. an additional hint "We already released this as a security fix in 2.93.1"? | 17:01:49 |
| leona | sorry don't have the energy today for writing a full issue and so | 17:02:02 |
| raitobezarius (DECT: 7248) | just so i understand well, we are talking about the release notes right? | 17:02:07 |
| raitobezarius (DECT: 7248) | no worries | 17:02:10 |
| raitobezarius (DECT: 7248) | we are not releasing every day thankfully :p | 17:02:16 |
| leona | yes | 17:02:17 |
| leona | like you could just write this sentence in the release notes for the items that were already released | 17:02:42 |
| leona | * like you could just write this sentence in the release notes for the parts that were already released | 17:02:46 |
| raitobezarius (DECT: 7248) | i agree with this | 17:03:07 |
| raitobezarius (DECT: 7248) | the RL were hastily merged unfortunately and we do not have a nice way to update them without point releases | 17:03:25 |
| raitobezarius (DECT: 7248) | it's documented as a known issue | 17:03:32 |
| raitobezarius (DECT: 7248) | ideally, this should happen automatically next time, but realistically, all we can do is to be better at running the releng process and having a checklist for that | 17:04:05 |
| raitobezarius (DECT: 7248) | PR welcome in checklist here: https://git.lix.systems/lix-project/lix/src/branch/main/releng :> | 17:04:26 |
 ShalokShalom | In reply to @522_:catgirl.cloud
does lix's readme have a mention of nixgl? yes | 17:53:19 |
 K900 | Uhh what | 17:57:25 |
| K900 | Where? | 17:57:43 |
 John Ericson | I am hoping we can agree on a new derivation format which fixes a number of issues | 18:23:44 |
| John Ericson | like versioning without central authority, and not stuff misc options in the environment variables, hijacking them | 18:24:15 |
| John Ericson | I am also interested in content-addressing of course, but it is fine if the format is merely forwards-compatible with such things | 18:24:55 |
| John Ericson | Edef has mentioned preference for ASN.1, that is fine with me | 18:25:17 |
| John Ericson | * Edef has mentioned a preference for ASN.1, that is fine with me | 18:25:42 |
 puck | mhm; i have some thoughts on this | 18:28:02 |
| puck | (but have had some trouble wording the ideas properly; i'll give it another try this weekend) | 18:29:26 |
 Martin Schwaighofer | Congrats on the release! 🥳
I think the breaking changes in the release notes might be missing the removal of the `ca-derivations` experimental feature. 🤔 | 18:32:18 |
| raitobezarius (DECT: 7248) | The last section does say that there's removal in 2.94 albeit it's worded bizzarely, the blog post does not make an explicit mention and I should add it probably, it was already deprecated in 2.93.1 so | 18:34:00 |
| John Ericson | you may wish to take a look at the DerivationOptions type I've added to Nix | 18:48:15 |
| John Ericson | I think our AST is getting pretty good shape, which can help narrow the scope for the choice of what is the concrete canonical message to sign | 18:49:15 |
 piegames | https://gerrit.lix.systems/c/lix/+/4605 🎉 | 18:51:37 |
| puck | well, i mean, i have Opinions on what is represented in a .drv, and really, what the usecase of a .drv is | 18:53:52 |
