On 4/30/07, Mario Rodriguez <biomates at telefonica.net> wrote:
> > I'd rather see some specific object type here. A list is very general.
>
> We don't need to re-invent the wheel. plot2d and plot3d have a syntax
> similar to this. I like the syntax object(ar1,arg2,...), but if there is
> a general agreement, I can change this to [object,ar1,arg2,...] in
> future versions. In this context, [explicit,expr,var,a,b] and
> [expr,var,a,b] could be considered equivalent.
Well, I'm hoping that we're going to get rid of the [foo, bar, baz] notation
in the plotting functions.
> > Just to throw out another idea (I haven't thought this through) consider this.
> > How about a notation which describes the set of points to be plotted?
> Sounds revolutionary from a CAS-point of view. This could be extended to
> other functions: {x: x^2+x+a=0} instead of algsys([x^2+x+a=0],[x]), etc.
> I think there are members in this list with better skills than me to
> tell us how far we can go in this direction.
I think I see what you mean here. I wouldn't mind something like
S : {x s.t. x^2 + x + a = 0};
elements (S);
to first represent the set implicitly and then to make an explicit
list of its elements.
In a sense the set-builder notation is the easy part;
the only difficulty is agreeing on the keywords or other symbols
and convincing the parser to recognize them. The hard part is
to actually do something interesting, e.g. apply algsys with
appropriate arguments or invent new algorithms. Be that as
it may, I'm pretty sure even a partial implementation of
set-builder stuff would be useful; we could get the syntax in
place and then fill in the computational backing piece by piece.
FWIW
Robert