Félix Saparelli
Félix Saparelli
I guess we could have urgent events (like a ctrl-c) interrupt the debounce loop.
I'm conflicted about this. I think it could add a lot of complexity for too little gain. For the record, this pattern is something we've encountered in cargo-watch and the...
Going to attempt something approaching this in 1.12.0 related to #101 and #132
This is for notifications via winrt-notifications.
Not directly at the moment, but that would be an interesting feature! Right now, the `$WATCHEXEC_*_PATH` variables (described in the readme) are your best bet: you can write a little...
This is a semi-popular request across both here and cargo-watch, but has too much complexity to be done in a dedicated fashion anytime soon. General idea would be something akin...
We also may want to provide more _detail_ on the events, up to the detail provided by `--print-events`, but in a machine-readable format. It would help for better understanding/parsing of...
That's [Known Issue Number 2](https://github.com/watchexec/watchexec/discussions/332)! You can work around for now with `--project-origin src`, or by starting watchexec in `src` and setting the cwd of the command with `--workdir ..`.
That's true, yes, we could likely get much better performance by optimising where file I/O happens. I'll have a specific look at this.
Yes, that's absolutely the case (well, kind of, it's complicated, there's actually two layers of figuring out ignore files and that's where the issues arise; happy to chat if you...