🚀 go-pugleaf

In article <danforth.828397346@Turing.Stanford.EDU>  
danforth@Turing.Stanford.EDU (Doug Danforth) writes:
> In <4jhm0c$ak8@ra.nrl.navy.mil> pitre@n5160d.nrl.navy.mil (Richard Pitre)  
writes:
> 
> ...
> >  ... The practical options
> >  are to include complex numbers or provide a mechanism for overloading
> >  the appropriate operations symbols. I think that the simplicity of the
> >  language would be almost completely unaffected by incorporating complex 
> >  numbers.
> 
>    Well, I can't help but wade in on this thread.  If we are really going
> to make the language look like mathematics 

Its not a matter of making it look like mathematics, its a matter of making
numbers act the same way everywhere in your code instead of having to
write complex code to deal with a distinction that
need not be there. 

> then I would request vector
> and matrix notation so that I can write:  A = B + C where, A, B, C

Its possible to support matrix operations and other types of operations in
favorite notations by incorporating operator overloading. I am not 
advocating for or against that. I'm saying that a trivial extension of the 
language would simplify a tremendous amount of numerical code by eliminating 
a lot of logic at the programmers level. The sole purpose of this extra logic
is to distinguish between one type of 
number and another where no distinction is necessary. This would open up 
a possibility of interest from a significant community of numerical 
programmers that would otherwise ignore Oberon. 
Oberon could use that support. I would like to see Oberon
flourish.

> are conformal tensor objects.  One also needs to be able
> to return such things from functions, e.g, B := f(A).  Unfortunately,
> we start to create something that looks very much like APL.
>    I suggest that the real solution resides in the next generation of
> computing devices where symbolic computations (and their related
> numeric calculations) are initiated with a pen.
> 

Languages that support symbolic computation have been around
for some time now. For good reasons it will be a while before they supplant
procedural languages to any significant degree. 



richard