Return rejected promise about broccoli-plugin HOT 6 CLOSED

Could this be a variation of isaacs/rimraf#72?

from broccoli-plugin.

I guess it is closely related, but why does broccoli tries to delete files before printing the error?
The issue I'm facing is the lost error message.

from broccoli-plugin.

why does broccoli tries to delete files before printing the error?

Ah, I see the problem - the cleanup happens before the error is reported, so the cleanup error "wins". It's a bit tricky to change this without silently swallowing cleanup errors, or adding a bunch of extra error handling code.

I'd like to suggest that it's actually acceptable to print the cleanup error rather than the build error. If a cleanup error happens, that's so fundamental an issue that it should usually be fixed first. I realize that this is happening a lot on Windows because of isaacs/rimraf#72, but then again perhaps one could say that Broccoli is simply broken on Windows until isaacs/rimraf#72 is fixed, and this is just one of many symptoms of this brokenness.

from broccoli-plugin.

But would not additional .catch before .finally suffice?

from broccoli-plugin.

I've thought about this some more, and I don't think the behavior should be changed. If there's a cleanup error, then that error should be reported.

from broccoli-plugin.

Fair enough. The isaacs/rimraf#72 is painful.

from broccoli-plugin.