[SciPy-user] scipy, matlab and NASTRAN

[Nils Wagner]

On Fri, 27 Feb 2009 11:33:54 -0800
  Matthew Brett <matthew.brett at gmail.com> wrote:
> Hi,
> 
>> +1.
>>
>> Anyway the source is available at
>> http://danial.org/op4/load_save_op4.tar.gz
> 
> Yes, but no use to us, because it's GPL.
> 
> Best,
> 
> Matthew

Maybe someone can convince Al Danial to release his code
under BSD license.

Nils

This is what I found on the mailing list:

If you are the copyright holder,
and if you have not granted exclusive rights
to another party, you can release you code
under as many different nonexclusive licenses
as you please as many times as you please.

If you created the code, you are the copyright holder.
If others helped create it, you need their "OK", unless 
they
assigned their copyrights to you.

So, for example, if you (the copyright holder)
released your code under the GPL and now wish
to release it under the new BSD license,
no problem. (And, bravo!)

Many people just hand out their code with a
note in the header stating the author and that
it is BSD licensed. If you want to be more
careful, the note should direct them to an
accompanying file (say, LICENSE.txt)
in which you copy this template at the *bottom* of
http://www.opensource.org/licenses/bsd-license.php
and replace <YEAR> and <OWNER>. (And actually,
this is short enough---2 paragraphs---that you can
put it in every file if you prefer.) Most people
nowadays leave out the no endorsement clause,
which proved pointless.

What does the BSD license do: it roughly says anyone can
use or modify the code without conditions except
keeping the copyright intact, and you incur no obligation
to them or to anyone else they might distribute to.
It is a good, simple license. If you leave out the
"endorsement clause", as I would, you end up with a
license that is essentially the same as the MIT
license. MIT is my favorite (for utter simplicity),
but for code destined for SciPy, the simplified BSD
(i.e., no endorsement clause) is the right choice.