[racket-dev] [plt] Push #25941: master branch updated

On 12/22/2012 05:04 PM, Matthias Felleisen wrote:
>
> On Dec 22, 2012, at 1:02 PM, Jens Axel Søgaard wrote:
>
>> According to
>>     https://www.cs.drexel.edu/~wan/publications/thesis.pdf
>> page 33 Gauss elimation turns onto a O(n^4) algorithm.
>> (Geddes mentions the same issue in relation to Gaussian
>> elimination over a general ring).
>
>
> Off topic question.
>
> Are you implementing GE as an instance of the GE over rings?
> If so, is it possible to formulate Unification and Type Inference
> for say STLC via your math library?
>
> Totally off topic. "No" is a fine answer.

No. But it's on-topic, so you get a slightly longer answer. :D

Because there's no generics support in Typed Racket, we're sticking to 
matrices of Real and Number for now. I'd love to have good typed 
generics for somewhat less esoteric things like polynomials, intervals, 
and the FFT, or for matrices with bigfloat elements.

Neil ⊥