You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on May 13, 2019. It is now read-only.
I'm not sure if this deserves an issue or if there's anything actionable, but anyway...
Lately, most of the failing tests I've found: nodejs/node#5342, nodejs/node#5339, nodejs/node#5129 were flaky because of the use (misuse?) of timers. It seems to me that a lot of times, the use of timers is not needed, as we can rely on the test runner timeout, and they become a source of flakiness. I could not agree more with this post from @Trottnodejs/node#5342 (comment). Should we set a position on this and maybe document it somewhere?
The text was updated successfully, but these errors were encountered:
I would love to get some consensus on whether arbitrary timers is (on the one hand) a valid way to provide a helpful error message to people running tests without the wrapper or (on the other hand) an anti-pattern that introduces maddening flakiness into our tests. (Spoiler alert: I think it's an anti-pattern.)
I'm not sure if this deserves an issue or if there's anything actionable, but anyway...
Lately, most of the failing tests I've found: nodejs/node#5342, nodejs/node#5339, nodejs/node#5129 were flaky because of the use (misuse?) of timers. It seems to me that a lot of times, the use of timers is not needed, as we can rely on the test runner timeout, and they become a source of flakiness. I could not agree more with this post from @Trott nodejs/node#5342 (comment). Should we set a position on this and maybe document it somewhere?
The text was updated successfully, but these errors were encountered: