From help-octave-request at bevo dot che dot wisc dot edu  Sat Dec  9 17:29:35 2000
Subject: Re: The future of Octave
From: j dot logsdon at lancaster dot ac dot uk
To: help-octave at bevo dot che dot wisc dot edu
Date: Sat, 9 Dec 2000 23:29:32 +0000 (GMT)

jwe has carried the load for far too long and has every right to call a
break.  He clearly deserves all our thanks.

It will in the long run be better for both John and Octave if he steps
back and lets someone else - or better a group - to take the lead but why
have things got to this level?

I don't really want to use too much bandwidth but while I still use Octave
on occasion, for most of my work I now use R - the statistical program.
Mainly this is because I do statistical work although the syntax is not as
natural as Octave's.

The comparison is interesting since R is an open source implementation of
the S language.  The comparable closed-source program is S-Plus which
curiously is also (now) owned by MathSoft.  Like Matlab, it is expensive
for a single user to buy (particularly on a Linux box). Are Mathsoft
profiteering?  Is too much of the licence fee going to sales and marketing
junkets chasing a rather difficult set of prospective purchasers rather
than slashing the cost?  The high (relative) cost of Matlab in a general
purpose environment means that there remains an important section of the
world community that needs a program like Octave (or Scilab or tlab or
...).

Since R is a program or environment for statistical calculations rather
than an operating system, this is a better comparison I think than the one
already mentioned of Linux.  

The R core team has managed to avoid overloading one member (actually it
started with two people anyway) and is spread world wide.  The program
continues apace and in a number of respects is faster and better than it's
commercial counterpart while missing some of the addons.  Most of the
libraries written (openly) for S-Plus have now been ported to R.

Why has the R approach worked and Octave one not?  Is it because there is
a clear target language, in which most libraries etc are written, rather
than a moving target where the target calls the shots?  Is it because
there is a better defined user-group (ie statisticians rather than anyone
who uses matrices from time to time)?  Is it because it is a higher level
language?

There are many other questions but the big problem is - what can be done
to bring Octave development closer to the R model?  Is it worth trying?
How do these groups develop?  This is really a problem I think for a
sociologist to answer but perhaps there are some thoughts on the list.

John




-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------