Simon Farnsworth

For now, most of our futures are `'static` because we still haven't fully moved off futures-0.1 (yes, I know, we need to finish migration), so spawning is not a problem....

@withoutboats How would I use `for_each_concurrent` to make the `do_select`s happen concurrently until I had a filled buffer, but *not* run multiple copies of `do_giant_work` concurrently? `do_giant_work` is quick to...

The pattern of concurrency I wanted was to have up to 5 `do_select` calls plus one `do_giant_work` call running concurrently. `do_giant_work` would spawn off more threads if needed, including `spawn_blocking`...

> No, this is not right. Buffered doesn't store the results of the the futures as they complete. That's exactly the problem this issue is describing, because that's why you...

I have an initial attempt at "proper" scoped tasks in https://github.com/farnz/tokio/commit/0ce475868061e8620fe52fb0d9e026f7db039ac8, but I need advice to continue. Specifically, I'm not clear on how `drop` should forcibly transition a task to...

> Sorry, scoped tasks cannot rely on `Drop` because you can use `mem::forget` to skip running the destructor. There is no way to write a scoped task API that can...

FTR, the reason we can't do crossbeam type scoping from an async context is that crossbeam blocks the thread that calls `scope` for the duration of the scoped threads execution....

This will fix issue #18

@morucci Once you've dealt with the CLA as per @facebook-github-bot's comments, this is ready to merge.