Adam C. Foltzer

Ah, I missed the implications of the changed type. This looks quite promising; my worry is whether we'll be able to make it work with the sparks-based scheduler. We'll have...

For this ticket, we are actually referring to safety in the Safe Haskell sense; with `Par` exported as a `MonadIO` instance, `runPar` amounts to `unsafePerformIO`. However, we want to keep...

This build failure is pretty puzzling to me, as I cannot duplicate it even with the exact same nightly compiler on Ubuntu 18.04: https://travis-ci.com/rust-lang/futures-rs/jobs/264248283

@taiki-e sorry for the delay on this reply. The reason it returns the combinator state is to not lose the bytes that have been buffered up to that point. If...

As requested, the GHCs where I've seen this include 7.8.3, 7.8.4, 7.10.2, and 7.10.3 for 64-bit Windows, 64-bit Mac, and both 32 and 64-bit Linux.

Hmm, looks like an error during dependency builds for the FreeBSD check

@seanmonstar We've had some success with a [revised set of "send" errors on the Fastly Compute platform](https://docs.rs/fastly/latest/fastly/http/request/enum.SendErrorCause.html) that are roughly based on [the `proxy-status` RFC](https://www.rfc-editor.org/rfc/rfc9209.html), though of course extended with...

You still do need a wrapper of some sort. However it should be open rather than closed, in the sense that if someone needs types other than what we've put...

I don't think GADTs alone are enough to get there. Consider this simplified example (`==` as a standin for `unify`), using GADTs in the way we might want: ``` haskell...

> But does this provide us a way of statically rejecting `(x == 1) && (x == True)` while statically accepting `x == 1`? I don't think so, but perhaps...