Hi Robert,
On Wed, 18 Jul 2007 08:11:48 -0600, Robert Dodier wrote:
> On 7/15/07, Neilen Marais <nmarais at sun.ac.za> wrote:
>
>> As discussed previously, the vect package defines "." as commutative,
>> thereby interfering with the use of "." for matrix multiplication (unless
>> I missed something in the mean time). I'm trying to make my code safe for
>> this situation.
>
> The present situation, in which the vect package redefines ".",
> is really a mess. If vect needs a commutative scalar product
> operator, then it should define a new one.
I wonder if it's not reasonable to define a different operator for matrix
multiplication, rather than the other way round? I.e. a ".." operator with
similar semantics to "^" and "^^"? Anyway, I agree that it is a good idea
to handle the "." operator in a more reasonable way, but I'd still be
interested in weather it's possible to write a wrapping function that
works as I described in my OP. I think it could be a nice metaphor for
other uses too.
> I think there was some work along those lines a while ago;
> I guess it's time to finish it. You proposed some changes to
> vect.mac and I think other people made suggestions too.
> I think changing vect.mac is very much preferable to trying
> to work around the current strange behavior.
I vaguely remember promising to spend some time on the topic back then,
but alas I haven't, and probably won't be able to soon. But it did seem
like some good ideas were being thrown around...
Regards
Neilen
>
> best,
> Robert Dodier