Proposal: Control.Concurrent.Async

[Simon Marlow]

On 15/06/2012 13:52, Twan van Laarhoven wrote:
> On 08/06/12 10:37, Simon Marlow wrote:
>> I'd like to add a higher-level concurrency API to the base package.
>
> How does this async package differ from monad-par, from a user's
> perspective? It seems that the latter is intended for pure computations
> and the implementation is more heavy weight. But it can be used for IO
> as well.

monad-par is for pure deterministic parallelism only, since it is based 
around

   runPar :: Par a -> a

So I'm slightly confused by your comment that it can be used for IO as 
well - what do you mean exactly?

I'd say monad-par is much more lightweight than async.  Perhaps not in 
terms of the amount of code in the library, but the overhead of 'fork' 
in monad-par is much less than the overhead of 'async'.

Incidentally, the IVar in monad-par is what I would call a "future" (in 
reference to the other suggestion that we use "Future" instead of "Async").

> The difference in API is essentially only that async lives in the IO
> monad, whereas monad-par has its own Par monad.

Not being in IO is a pretty big difference :)

> Could and/or should one be implemented in terms of the other?

We could implement monad-par in terms of Async, but I doubt it would 
make a lot of difference.  The reverse is obviously not possible.

Cheers,
	Simon