Re: [Maxima-discuss] proposal to include lapack in memory image,

[Robert D. <rob...@gm...>]

On 2014-09-17, Raymond Toy <toy...@gm...> wrote:

> I prefer option 1 because it runs on any CL, and the amount of work to
> maintain and test is fairly minimal. And we're not trying to win any
> speed contests here, so functionality trumps speed for me.

Same here.

> Oh, one other thing.  With maxima's bigfloat numeric package and the
> lapack sources in lisp, it is straightforward to convert the code so
> that lapack can operate on arrays of bigfloats.  It won't be fast, but
> that's something lapack doesn't support. (But maybe mpfr does?)

Yes, that's a worthy goal.

best,

Robert Dodier