| 18 Jul 2025 |
 dramforever | if it's cross it uses pkg-config for build and PKG_CONFIG_underscored_triple for host? | 20:57:31 |
| dramforever | really? | 20:57:40 |
| dramforever | i thought it replaces '-' with '_' | 20:57:48 |
| dramforever | otherwise why would they call it target_u | 20:57:57 |
| dramforever | oh wait it checks both | 20:58:23 |
| dramforever | ah yes my environment variable called foo-bar | 20:58:44 |
| dramforever | this is definitely not a usability nightmare | 20:58:53 |
| dramforever | anyway underscored should be fine | 20:59:11 |
| dramforever | we do that in nixpkgs too right? | 20:59:16 |
 emily | no | 20:59:21 |
| emily | it has no notion of build or host | 20:59:23 |
| emily | AFAICT | 20:59:27 |
| emily | like | 20:59:30 |
| emily | as in | 20:59:30 |
| emily | the API will always look for host platform deps | 20:59:37 |
| dramforever | oh so you just can't get build pkg-config | 20:59:42 |
| emily | yeah, only via dev deps | 20:59:52 |
| emily | it just will adjust whether it looks at HOST_* or TARGET_* based on whether the current build is cross | 21:00:02 |
| emily | so you can mix multiple pkg-config in a single cargo build | 21:00:10 |
| emily | but each build script only sees one platform for pkg-config purposes | 21:00:20 |
| emily | anyway it's bad and I hate it | 21:00:37 |
| dramforever | this is cursed but i can't think of anything off the top of my head that really needs the other | 21:01:32 |
| dramforever | so whatever | 21:01:35 |
| emily | well | 21:03:41 |
| emily | the package we are trying to fix is sort of one of them | 21:03:47 |
| emily | kind of? | 21:03:48 |
| emily | it's weird | 21:03:49 |
| emily | I really think the GNU target-specific cross model ruined everything forever. | 21:04:24 |
| emily | if Plan 9 had won everyone would be much less confused about cross. | 21:04:36 |
| dramforever | it always bothers me how gnu toolchains just does this | 21:05:32 |
