| 25 Sep 2024 |
 Infinidoge 🏳️⚧️ | * When you've gone to the level of inventing Javaagent-based raw bytecode manipulation tools, the base game stops being an obstacle lmao | 03:28:40 |
 emily | I like to imagine a Nix-using business trying to explain to auditors why their JDK comes from github.com/MinecraftModTeam | 03:29:16 |
 Philip Taron (UTC-8) | "You know what was the original SBOM? A minecraft seed. And you know it's worth seeing what it grows into." | 03:43:03 |
|  luna-null changed their display name from luna-null to Autumn. | 06:39:21 |
| emily | finally done https://github.com/NixOS/nixpkgs/pull/344544 | 23:06:38 |
| 26 Sep 2024 |
|  Fabián Heredia set a profile picture. | 01:16:08 |
| Infinidoge 🏳️⚧️ | In reply to @philiptaron:matrix.org
"You know what was the original SBOM? A minecraft seed. And you know it's worth seeing what it grows into." As in Software Bill of Materials, or is there another acronym I don't know about? | 19:42:08 |
| 30 Sep 2024 |
|  Frédéric Christ 🌴 16.09. - 30.09. changed their display name from Frédéric Christ 🌴 16.09. - 30.09. to Frédéric Christ. | 08:50:21 |
| 1 Oct 2024 |
|  -_o joined the room. | 21:01:20 |
| 2 Oct 2024 |
|  vandycarlos left the room. | 12:12:19 |
| 4 Oct 2024 |
|  @terrorjack:matrix.terrorjack.com joined the room. | 17:41:47 |
| 7 Oct 2024 |
|  memegames99 removed their profile picture. | 00:05:41 |
| memegames99 removed their display name memegames99. | 00:05:51 |
| memegames99 left the room. | 00:06:09 |
| emily | I'm working on deduplicating the OpenJDK/OpenJFX expressions. | 00:49:03 |
| emily | it's not pretty. the variance between versions is pretty cursed, as is often the case with these copy-paste situations | 00:49:22 |
| emily | but I have something that reproduces the exact derivation hashes of all the current versions from one file | 00:49:33 |
| Infinidoge 🏳️⚧️ | Very nice | 00:49:45 |
| emily | it needs clean-up to not have a bunch of hardcoded stuff, but I'll put a PR up after that | 00:49:48 |
| emily | I won't do any further refactoring right now, since I think it's best to have a zero-rebuild PR that encodes the exact ways in which the version-specific logic is messed up and then work from that | 00:50:08 |
| Infinidoge 🏳️⚧️ | That variance is why I didn't attempt to clean up the files myself, I just didn't have the time to sit and work through it lol | 00:50:15 |
| Infinidoge 🏳️⚧️ | Yeah, that's for the best | 00:50:21 |
| emily | what I did was start with 23.nix, and then diff 21.nix → 23.nix (after running both through nixfmt to account for formatting differences). everything that changed is something that needs a conditional | 00:50:50 |
| emily | then I just repeated that with 17.nix → 21.nix, etc. | 00:50:57 |
| emily | JDK 8 was the worst, since a bunch of things are just different from every subsequent version. | 00:51:13 |
| emily | after that I may nerd-snipe myself into clean-up → Darwin support → looking into getting rid of one of the two ~identical binary OpenJDK distributions | 00:52:12 |
| emily | I'm referring to the Temurin build scripts a lot while working on this. I'm wondering if we shouldn't consider simply using them to build OpenJDK if possible. | 14:56:59 |
| emily | they handle a lot of the tangled per-version logic for us and probably have settings that more reliably make sense. | 14:57:19 |
| emily | they also already have Darwin support. | 14:58:24 |
| emily | turns out that OpenJFK and OpenJFX 11 support cycles aren't aligned and I missed a >year-old WebKit 🫠 https://github.com/NixOS/nixpkgs/pull/347149 | 18:54:11 |
