Stream: contributing

Topic: no output when executing `roc subcommand`


view this post on Zulip Anton (Oct 24 2025 at 18:07):

With the zig 15 migration I forgot to flush after writing in a bunch of places. This may cause you to see nothing when executing roc check. I'm working on a fix in #8329 but I still need to clean that up. I just wanted to give you all a heads up in case you hit this issue.

view this post on Zulip Richard Feldman (Oct 24 2025 at 18:24):

I fixed that on a branch

view this post on Zulip Richard Feldman (Oct 24 2025 at 18:24):

or at least some of them, might have missed some! :smile:

view this post on Zulip Isaac Van Doren (Oct 24 2025 at 20:18):

I haven’t checked recently, but there may be some places where std.proccess.exit(1) is being called and preventing writers from being flushed also. It would probably be good to move to only calling that at the very top level and propagating errors up so that everything gets flushed

view this post on Zulip Richard Feldman (Oct 24 2025 at 21:05):

yeah we shouldn't be doing that anyway because it defeats the "nonblocking compilation" philosophy of making it so you can always run the program if you want, no matter what errors happen during the build

view this post on Zulip Anton (Oct 27 2025 at 18:51):

Isaac Van Doren said:

It would probably be good to move to only calling that at the very top level and propagating errors up so that everything gets flushed

I have started on this.


Last updated: Nov 09 2025 at 12:14 UTC