octave-maintainers
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Wrapper for vectorized math libraries


From: Jordi Gutiérrez Hermoso
Subject: Re: Wrapper for vectorized math libraries
Date: Sat, 1 Oct 2011 11:52:13 -0500

On 1 October 2011 11:26, Dan Davis <address@hidden> wrote:
> Thus, I pose the question, before I spend too much time hacking for
> the sake of hacking, if I could give Octave the capability to
> utilize these closed source libraries' vectorized transcendental
> function calls without having to distribute them with Octave, and
> thus, as far as I can tell, not violating any license, would that
> capability be desired?

At least on my part, no. You may keep such a script for yourself if
you wish, but we cannot accept it in Octave, because it violates the
GNU coding standards:

    http://www.gnu.org/prep/standards/standards.html#References

The GPL should not be an obstacle to be worked around with hacks. "If
I distribute it this way, can I avoid the copyleft?" seems to me like
a misguided question. The real obstacle is that there is no free
library that does what you need, so the real solution is to find a way
to get a free library to do what you need. I recommend you don't start
from scratch. I wasn't able to follow all of your discussion, but as I
understand it, you would like a free library that has vectorised calls
for some special functions. The GSL is an example of a free library
that can do special function, so why don't you work on trying to build
vectorisation into the GSL instead? I find its source to be very
readable C, and perhaps you will enjoy reading it too.

I think that is a much better long-term investment of your time. I
understand you may not want to do it and would prefer a simple legal
hack for a perceived smaller problem. I, however, see a larger problem
and I would prefer to see a larger solution to it. I hope you decide
to take this route instead.

- Jordi G. H.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]