Untyped Scheme should be built on Typed Scheme? WAS: Re: [plt-scheme] macro question

On Sun, Dec 21, 2008 at 7:31 PM, Grant Rettke <grettke at acm.org> wrote:
> Hi folks,
>
> On Sun, Dec 14, 2008 at 7:36 PM, Grant Rettke <grettke at acm.org> wrote:
>> Hi Matthias,
>>
>> On Tue, Jun 10, 2008 at 10:22 AM, Matthias Felleisen
>> <matthias at ccs.neu.edu> wrote:
>>> On Jun 10, 2008, at 11:40 AM, hendrik at topoi.pooq.com wrote:
>>>> In my opinion, untyped Scheme needs to be built on top of typed Scheme,
>>>> not the other way around.  But until this revolution happens, I'm
>>>> happy to use it they way it is.
>>>
>>> NSF wasn't willing to fund a time machine, even when I promised I'd add cold
>>> fusion.
>>
>> You were joking around here, but is there any truth to it?
>>
>> Theoretically if you could start over, would you implement Untyped
>> Scheme on top of a Typed Scheme?
>
> Was this a dumb question or did no one reply because I only asked Matthias?

I think the current layering is the right one.  This is because in
general, it's better to build more complex systems from simpler ones,
and the untyped system is simpler.

Thanks,
-- 
sam th
samth at ccs.neu.edu