@bjorn3 I believe cargo-bisect-rustc#83 should fix the issue you've reported
didn't have time to properly test it, if you or someone can test it please go ahead, otherwise I'll try to properly test later
I can probably try Saturday. Can you please @ mention me at the PR, so that the gh notification acts as a reminder for me?
@bjorn3 if Santiago's changes address this, please let me know and we will push the fix with other changes that are in master over the weekend.
@bjorn3 just pushed a different fix for this one
it's a tiny bit ugly
let me know how can I test exactly what you're testing
Just pushed a commit to cargo-bisect-rustc-bot-jobs to test it: https://github.com/bjorn3/cargo-bisect-rustc-bot-jobs/runs/664142625
searched toolchains nightly-2020-04-25 through nightly-2020-04-25
Regression in nightly-2020-04-08
@bjorn3 ouch this is another issue and I see what that is, if it's easy for you to lunch tests I may keep asking you to try otherwise if you can show me how to reproduce this would be nice too :)
@bjorn3 pushed another attempt to fix it, test it again or let me know a way of testing this
this latest fix works, just tested your repo and reported back on the issue
@Chris Simpkins you around?
we clearly need a test suite here :joy:
anyway, I've addressed the new problems
Yep! I am here. Definitely agree with the test suite issue and commented on your PR thread. I am up for diving in on a refactor with test coverage in mind. Problem on my end is time for a significant refactor right now
@pnkfelix mentioned that he began work on testing and IIRC this is in a branch of his fork of the repository. Perhaps we can begin there and see what direction he is heading with this? If you'd like to get together to run through how to refactor with other SH's in the project, count me in.
yeah same happens to me, the main problem is finding time :)
but :+1: on what you've said
Posted this in the "Testing" thread: https://rust-lang.zulipchat.com/#narrow/stream/217417-t-compiler.2Fcargo-bisect-rustc/topic/Testing/near/197299207
I think the PR should be ready to be merged now
@Chris Simpkins given that there are a bunch of "unrelated" things in the PR, I'd recommend reviewing commit by commit
Will take a look this evening.
@bjorn3 any concerns about merging Santiago's work to address your bisect-rustc issue? It looks like this is sorted out and should be fixed as of his latest commit. We are having a more general conversation about how to approach the reporting and it would be useful to hear from you what elements in the report are useful/required for your GH Actions based tool. Separate issue, but please let us know if there is a way to better support your approach.
@Chris Simpkins Merging is fine.
The main problem with my bot right now is that github doesn't allow me to access github actions logs using a personal access token. I think the documentation implies that you have to create a github app to access it.
This fix was merged. We will push a release this week.