On Mon, Feb 23, 2015 at 11:13 AM, Keean Schupke <[email protected]> wrote:

> On 23 February 2015 at 19:01, Jonathan S. Shapiro <[email protected]>
> wrote:
>
>>
>> Only where we know the closures can be stack-allocated. This requires
>> knowing that they won't be captured, and I haven't thought that through
>> hard enough.
>>
>>
> See, this is where your feedback gets a bit weird, and I can't help
> thinking one of us is misunderstanding, so can we slow down:
>
> If I define:
>
> def f x = \y . x + y
>
> and I can then use this as:
>
> ((f 1) 2
>
> All I am suggesting is that you can write this:
>
> (f 1 2)
>
> And have the compiler 'specialise' the call to:
>
> ((f 1) 2)
>

I agree. I'm not quite sure what I was thinking about.

Sorry for mis-leading/mis-reading you.


shap
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to