make_random_state broken for Maxima installed with Ecl

On Fri, Aug 29, 2008 at 09:51:43AM -0600, Robert Dodier wrote:
> On Fri, Aug 29, 2008 at 9:47 AM, Oliver Kullmann
> <O.Kullmann at swansea.ac.uk> wrote:
> 
> > I have now the problem that I use make_random_state(0)
> > at quite a few places.
> >
> > So it seems I need to assemble my own Maxima package:
> >  - the module "maxima" of the cvs repository is to be used?
> >  - once I see on "Index of /maxima/src"
> > http://maxima.cvs.sourceforge.net/maxima/maxima/src/?sortby=date#dirlist
> >   that the changes I'm looking for appeared, I have to
> >   download the module "maxima", go into the 5.16.3 package
> >   and replace all related entries there with the downloaded
> >   ones?
> 
> That seems like a lot of work. Why not replace 0 by 1?
> Is there something special about the state generated from 0?
>

It's about guidance: Setting the state happens in testfunctions 
(in those which test randomised functions), and it seems a little
nicer to use the "first" possible seed. Now if 0 is available, but
1 is used, this looks a bit careless (or if at some place 0 is
used, at others 1).

("Guidance" relates to the fact, that my testfunctions are 
ordinary functions (not hidden), only "higher order", and that the code in
my library should have "model character".)

Oliver
 
> Robert Dodier

-- 
Dr. Oliver Kullmann
Computer Science Department
Swansea University
Faraday Building, Singleton Park
Swansea SA2 8PP, UK
http://cs.swan.ac.uk/~csoliver/