| 28 Sep 2022 |
 Okno | >- PR is a fix: you explain exactly what you're fixing in a few words / sentences ( a crash occurs when ... This PR fixes it) -> screenshots if you have visual cues of it
How would you automate that, pull request template ? | 07:23:47 |
 pj | you can add codecov without having tests? | 07:24:14 |
 nheuillet | Oh yeah I see what you mean | 07:24:35 |
| nheuillet | Yeah that's where is starts being annoying, ideally we would mock and test it still but I don't think we need to that for now | 07:24:57 |
| nheuillet | ^ | 07:25:10 |
| Okno | For what its worth I have a crate interacting with git repository and I had to write this to get the test correct : https://docs.rs/sealed_test/latest/sealed_test/ | 07:25:12 |
| nheuillet | Most likely yeah, up sure there are ways to better enforce it but PR templates is a good start | 07:26:06 |
| Okno | 
Download screenshot-2022-09-28T093633.png | 07:37:05 |
| Okno | There is room for improvement 😆 | 07:37:20 |
| nheuillet | Lapce rpc should be too in some way | 07:40:27 |
| nheuillet | But yeah good coverage on core would be amazing | 07:40:28 |
| Okno | just sent a PR adding coverage to the CI, I guess a maintainer needs to authorize codecov github app | 07:51:19 |
| pj | why remove tests though | 07:52:04 |
| Okno | cargo llvm-cov run the test | 07:52:29 |
| Okno | no need to run twice | 07:52:36 |
| pj | then please note that in PR | 07:52:41 |
| nheuillet | btw what's the state of code coverage on rust? Last time I checked was two years ago and it seemed like a bother, how's the speed / easiness now? | 07:55:43 |
| Okno | It has improved a lot | 07:56:16 |
| Okno | compilation is still a bit longer than for normal test execution but I guess this is not something we really care about for CI builds | 07:57:01 |
| nheuillet | Yeah that's fair | 07:58:07 |
|  tmp joined the room. | 08:17:19 |
 MinusGix | 
Download unknown.png | 08:21:52 |
| MinusGix | seems like it llvm-cov triggers for derives, which might sometimes be useful but I wonder if we could disable that | 08:21:53 |
| MinusGix | Unfortunately, I don't seem to see a nice method on the cargo-llvm-cov repo | 08:37:44 |
| Okno | I don't know if that's an issue, you can reasonably hit 80% coverage even with derive impl picked by cargo llvm is think. | 09:03:39 |
| Okno | + coverage is not an absolute measure, only the diff matters | 09:11:58 |
|  inagun joined the room. | 09:29:06 |
| Okno | I am writting some unit tests in lapce-core, starting with word.rs, btw I think it's better if those are written by someone less familiar with the codebase than you guys. Less knowledge, less assumptions | 09:35:58 |
| nheuillet | Not the hero we deserve, but the hero we need | 09:37:01 |
| nheuillet | jokes aside, thank you so much | 09:37:09 |
