[Scilab-Dev] Future change in the Scilab memory representation.

[Sylvestre Ledru]

Le mercredi 04 mars 2009 à 11:30 -0400, Jonathan Blanchard a écrit :
> Waa cool no more stack?
Yes, it is going to be a pleasure ;)

> How is it going to work? Garbage collector, custom memory allocation?
Custom memory allocation but we keep in mind the idea of a GC.

> Also does this mean that we could allocate more memory than the amount
> of physical memory?
It is the goal ! :)

> Anyways Thank you for the answer.
You are welcome.

Sylvestre

> Jonathan Blanchard
> 
> 
> 
> On Wed, Mar 4, 2009 at 11:22 AM, Sylvestre Ledru
> <sylvestre.ledru at scilab.org> wrote:
> > Le vendredi 20 février 2009 à 13:14 -0400, Jonathan Blanchard a écrit :
> >> Hi,
> > Sorry for the delay.
> >
> >> As I said before I'm currently working on a toolbox enabling
> >> multiprecision matrix calculation using mpfr/gmp.
> > Excellent! I am looking forward to see that.
> >
> >> Straight to the point, I've thought about the memory representation of
> >> matrix for a while now and initially decided to use the same
> >> column-major representation as Scilab use.
> >>
> >> From what I recall this is the Fortran way of doing things. Is there
> >> plan to change this in the future?
> > The memory data structure is going to remain the same.
> > However, the stack is going to disappear in the version 6.
> >
> > Sylvestre
> >
> >
> >