[plt-dev] Inexact integers

That's one application area I had in mind, the other one is embeddings.


On May 25, 2010, at 9:49 AM, Shriram Krishnamurthi wrote:

> Isn't the numeric tower just another matter of language semantics,  
> after all?

(I sure don't know what 'just semantics' means from a language person.  
Sounds like someone else :-)




>
> [It sure would be nice to provide a "JavaScript numbers" language
> level in DrScheme, so people can test under browser numbers and make
> sure those are sufficient for their programs.  If they are, Moby can
> generate much simpler and vastly faster code.]
>
> Shriram
>
> On Tue, May 25, 2010 at 9:46 AM, Matthias Felleisen
> <matthias at ccs.neu.edu> wrote:
>>
>> What you're pointing out with these obvious claims is that the  
>> proposal
>> deserves some deliberation.
>>
>> (I do sometimes wonder whether we should parameterize the language  
>> over the
>> numeric tower. It is my understanding that the tower consumers a  
>> large
>> amount of 'footprint' and I bet there are applications where machine
>> integers are just enough. But that's just a dream probably.)
>>
>>
>>
>>
>> On May 25, 2010, at 3:31 AM, Michael Sperber wrote:
>>>>
>>>> Well, but the decision to have (integer? -17.0) => #t is rooted  
>>>> in the
>>>
>>> design of the numerical tower.  In particular, I would think it has
>>> implications for:
>>>
>>> (= -17.0 -17) => ?
>>> (rational? -17.0) => ?
>>>
>>> etc.
>>>
>>> --
>>> Cheers =8-} Mike
>>> Friede, Völkerverständigung und überhaupt blabla
>>> _________________________________________________
>>>  For list-related administrative tasks:
>>>  http://list.cs.brown.edu/mailman/listinfo/plt-dev
>>
>> _________________________________________________
>>  For list-related administrative tasks:
>>  http://list.cs.brown.edu/mailman/listinfo/plt-dev
>>