Re: Test web-site of new octave-forge web-site including all packages

[Joe Koski]

on 10/10/06 3:40 PM, John W. Eaton at address@hidden wrote:

> On 10-Oct-2006, Joe Koski wrote:
> 
> | As one of the few who actually uses gcvspl with it's octave interface, I
> | would hate to see it go, but I understand the dilemma. For what it is worth,
> | gcvspl.f is available for download on from www.netlib.org. How about leaving
> | the octave interface available (in the online help files?), but directing
> | those that want to use it to NetLib?
> 
> Perhaps you did not see the later explanation of why this is a
> problem.
> 
> | What are the license
> | requirements for use of NetLib software anyway?
> 
> AFAIK, there is no uniform license requirement for code to be included
> in the netlib collection..
> 
> | Could it be that gcvspl
> | licensing has changed since it entered octave-forge?
> 
> Yes, it is possible.  Does anyone know precisely what the license is?
> 
> jwe

John,

I made one try to find H. J. Woltring, the author of the gcvspl code. I
received the following reply:
===================
Dear Dr. Koski,
I regret to tell you that Dr. Woltring died long ago in a traffic
accident. At the time at which this happened Dr. Woltring was not
affiliated with my University anymore. I have no idea where his computer
programs went. Neither have do I have any written material other than
what is published in Journals. Good luck to you; sincerely Rik Huiskes,
Ph.D. 
====================
So I guess that gcvspl is a lost cause. I'll keep a copy with the octave
bindings for my use.

One thought on the new octave-forge package system. Apple seems to take
great delight in rearranging the furniture in the Xcode Developer Tools
regularly, with the result that they break mkoctfile in binary distributions
built with an older OS X. This means that the focus for folks like HPC and
Fink will be to keep mkoctfile working, and let Mac users get their own
octave-forge packages as they need them. Correct?

Joe