matthew healy

Hey Ben - that's totally fair, thanks for letting me know! I'll take a look at cats MTL instead. 😁

As of 0.3.0, this is how `match` works, so I think we can close this issue.

Hi folks! I ran into this exact problem today & ended up reimplementing `AfterCommitAsyncDispatcher` to instead delegate to a child dispatcher. If the change to the scheduler API discussed here...

Thank you so much for replying, and apologies for the long delay in getting back to you. I think your suggestion of using `subscribe_to_all_events` makes a lot of sense. I'm...

Probably worth noting that I discovered this while rewriting [this test](https://github.com/tweag/nickel/blob/master/tests/integration/stdlib_arrays_fail.rs#L13) which already parenthesises the negative number, so I guess this was known about at some point.

After discussing this with @vkleen on Slack, this isn't _really_ a bug. It's possible we could tweak the parser to make it so `f x -1` parses as a number...

> (what would be the conditions, by the way?) At least in my head, I'd like `f x -1` to parse as `f x (-1)` and `f x - 1`...

> At the very least I think this assertion needs to be supported by facts (could be measured in other programming languages). This is very reasonable! I was mostly just...

Yeah, I agree that could be a reasonable first step. Though since stdlib releases are currently tightly coupled to language releases I wonder if we should explore the design space...

@jneem Yeah, as your edit says, the difference between the case in the issue and yours is the "found" type. If the error in the issue said that `Number` had...