Conversation
|
|
||
| req.on('error', (err) => { | ||
|
|
||
| expect(err).to.not.exist(); |
There was a problem hiding this comment.
In some (flaky) cases this error was being triggered late and causing the next test to fail.
There was a problem hiding this comment.
Yeah, lab does not bind the expect to a specific test, so it can only assume that errors are caused by the current test.
kanongil
left a comment
There was a problem hiding this comment.
Mostly nits, but the "handles race condition between equal client and server timeouts" rewrite seems like it is not doing the same thing.
|
|
||
| req.on('error', (err) => { | ||
|
|
||
| expect(err).to.not.exist(); |
There was a problem hiding this comment.
Yeah, lab does not bind the expect to a specific test, so it can only assume that errors are caused by the current test.
|
@kanongil I appreciate the review, and I've incorporated your feedback. |
|
I appreciate all the careful review 👍 |
|
For what it's worth I still see these on master I don't have the bandwidth to dive into this atm but if I can help by running some tests or something let me know. (for quicker response dm me on slack) |
There are a few issues to fix for failures in citgm:
CC @AdriVanHoudt I know in nodejs/citgm#436 you were able to reproduce the macos issues. If you have time and would be open to trying out these changes it would be very much appreciated.