Bas van Dijk

There's some common code that can be abstracted: ``` haskell concurrently :: IO a -> IO b -> IO (a,b) concurrently left right = do mv = putMVar mv l...

> So as it stands I can't accept your version of concurrently: although I agree the semantics is reasonable, it doesn't match the existing specification of concurrently, and we can't...

What if we just change the spec to: ``` haskell concurrently left right = withAsync left $ \a -> withAsync right $ \b -> link2 a b waitBoth a b...

Hi Simon, what do you think about my patch?

Yes, I also [do that](https://github.com/simonmar/async/pull/21/files#diff-54890a66957e2efe3f5ceaa39499ff45R237) in #21.

Thanks for the update Julien! Since any CODEOWNER (person or team) is automatically requested for review by GitHub we're considering adding the Mergify condition: ``` - "#approved-reviews-by=#review-requested" ``` to approximate...

I'm getting a `The new Mergify configuration is invalid` when adding the `#approved-reviews-by=#review-requested` condition. So I guess that's not supported by Mergify.

The bug is here: https://github.com/bors-ng/bors-ng/blob/524eff45d09e0ec052346ae88d06e8c3f92a1003/lib/github/parse_owner.ex#L61

+1

I don't have an idea yet why the [stressTest for RWLocks](https://github.com/basvandijk/concurrent-extra/blob/f75f1b98f015b8074614fd9b333793239cd07e0f/Control/Concurrent/ReadWriteLock/Test.hs#L99) is failing. Maybe the timeout of `500 * a_moment` (2.5s) is too small. In any case, I'm not happy...