| 31 Jan 2023 |
@qbit:tapenet.org | you could do something like in pkgs/misc/logging/beats/7.x.nix | 15:29:57 |
@qbit:tapenet.org | is the go source in the kitty repo? | 15:30:30 |
@qbit:tapenet.org | oh, sorry - i though tbeats was building some other non-go package, looks like they are all go | 15:30:57 |
@adam:valkor.net | looks like yes https://github.com/kovidgoyal/kitty/tree/master/tools/cli | 15:31:21 |
@qbit:tapenet.org | maybe just have the go 1.19 compiler available and use the existing derivation ? | 15:33:37 |
@qbit:tapenet.org | it's not immediately clear where it's actually building the go stuff (nothing obvious in the makefile) | 15:34:10 |
@adam:valkor.net | the modules aren't vendored though, so it looks like it's failing on pulling them | 15:34:13 |
@qbit:tapenet.org | ohhh, gross | 15:34:25 |
@qbit:tapenet.org | maybe have a let theSource = fetchFromGitHub {} and use that in two different derivations? | 15:35:55 |
@qbit:tapenet.org | not sure how you'd combine it back into one though.. maybe makeWrapper | 15:36:13 |
@adam:valkor.net | ugh. https://github.com/kovidgoyal/kitty/blob/5a997a5f7a3f28dec1f9675cd434ae6d2c92837a/setup.py#L899 | 15:38:21 |
@qbit:tapenet.org | gross | 15:39:05 |
| * @qbit:tapenet.org has flashbacks of node-gyp | 15:39:16 |
Sandro 🐧 | you probably want to build that in an extra derivation and copy over to the python stuff | 17:14:46 |
@adam:valkor.net | Yeah, that's what I'm working on now | 18:20:04 |
@adam:valkor.net | ok, here's another question. How can I deal with the improper module name in the go.mod? https://github.com/kovidgoyal/kitty/blob/5a997a5f7a3f28dec1f9675cd434ae6d2c92837a/go.mod#L1 | 20:39:15 |
@adam:valkor.net | Patch the entire project? Or....? | 20:39:31 |
@qbit:tapenet.org | uuugh | 20:41:11 |
@adam:valkor.net | yeah, i will report it upstream as well | 20:41:37 |
@qbit:tapenet.org | maybe you could do the vendor dance initially (to get the vendor dir), then disable go modules and do the build.. | 20:41:55 |
@qbit:tapenet.org | maybe.. i am not sure our buildGo stuff will permit that | 20:42:06 |
@qbit:tapenet.org | probably be easier to patch all the go files though | 20:42:57 |
@adam:valkor.net | I don't think that's actually my problem. Sigh. | 20:55:56 |
@adam:valkor.net | python is being used to generate go code | 20:56:23 |
@adam:valkor.net | Though this worked well: https://github.com/sirkon/go-imports-rename :) | 20:57:20 |
@qbit:tapenet.org | yay, the sourcehut GOPRIVATE stuff is not gonna happen | 21:36:58 |
@qbit:tapenet.org | Update 2023-01-31: Russ Cox of the Go team reached out to us to address this problem. After some discussion, an acceptable plan was worked out. The Go team is working on deploying an update to the “go” tool to add a -reuse flag, which should substantially reduce the traffic generated by this system for all users of Go. | 21:37:03 |
@qbit:tapenet.org | https://sourcehut.org/blog/2023-01-09-gomodulemirror/ | 21:37:09 |
| 1 Feb 2023 |
Sandro 🐧 | In reply to @adam:valkor.net ok, here's another question. How can I deal with the improper module name in the go.mod? https://github.com/kovidgoyal/kitty/blob/5a997a5f7a3f28dec1f9675cd434ae6d2c92837a/go.mod#L1 haha, good joke | 13:40:02 |
Sandro 🐧 | clone the repo and build in place but our tooling could break | 13:40:24 |