[ale] anyone running Octave ?

Mills, John M. Mills.J at ems-t.com
Wed Jan 4 09:20:09 EST 2006


ALErs - 

My initial experiences with Scilab and Scicos have been very good. See
notes below.

-----Original Message-----
From: ale-bounces at ale.org [mailto:ale-bounces at ale.org] On Behalf Of
To: ale at ale.org
James P. Kinney III
Sent: Tuesday, January 03, 2006 11:27 AM
To: Atlanta Linux Enthusiasts
Subject: Re: [ale] anyone running Octave ?

On Tue, 2006-01-03 at 10:41 -0500, Mills, John M. wrote:

> The reason [I didn't use it] is that I was porting a MATLAB model and
> used Scilab instead. Depending on what you're doing, you may want to
> look into Scilab and its 'Scicos' tool (www.scilab.org).

JK> Thanks for the pointers!! Wish I had been able to use this in grad
JK> school. Scicos looks _very_ exciting.

Yes, well, me too ... but that was shortly after the discovery of fire
and well before the age of solid, free, user-supported software. More
seriously, I'm a relatively inexperienced user of these tools, but my
experience with Scilab and Scicos have been very good and both seem to
be in a period of lively growth in applications and extensions. 

I haven't tried Scilab's symbolic math tools (either free or purchased),
so I can't speak for them.

Scilab/Scicos doesn't have the staggering backlog of developed
applications and 3rd-party tools of MATLAB, but doesn't have the
staggering $15K+ price tag, either [for the set of toolboxes to support
our consultant's models]. That means anyone with interest can test and
follow-up your work: a big "plus" in my opinion. Favorable as to speed
on my [relatively simple] models, though the 'drag-and-drop' environment
is less polished. Scilab may be pulling ahead in using XML, in fact. I
think the integration with other languages is a bit easier if you are
doing it, but again there aren't so many pre-built applications around.

I've used Mathcad's symbolic math ability: it's great for development
and explaining "on the fly." Scilab doesn't have the "notepad" character
of Mathcad, but it doesn't have Mathcad's treacherous computation
dependence on page layout, backwards incompatibility with recent version
changes, nor the unique "poision and save" tool that killed more than a
few of my Mathcad models (else required that I pick through XML looking
for malformed blocks)! Altogether I use Mathcad only when I _really_
need to. Naturally the $multiK license fee is a factor, too!

Bottom line: the only ones of these tool accessible to me on my own
nickle are Octave and Scilab/Scicos/sibs. Fortunately those seem quite
effective - if sometimes a bit short of polish.

 - Mills




More information about the Ale mailing list