Mailing-List: contact legal-discuss-help@apache.org; run by ezmlm
Precedence: bulk
Message-ID: <42291F90.7030503@apache.org>
Date: Fri, 04 Mar 2005 21:55:12 -0500
From: Phil Steitz <psteitz@apache.org>
User-Agent: Mozilla Thunderbird 1.0RC1 (X11/20041201)
MIME-Version: 1.0
To: legal-discuss@apache.org
Subject: Re: Public Domain Software
References: <32D5845A745BFB429CBDBADA57CD41AF0CDD15D9@ussjex01.amer.bea.com>
 <20050111155744.O7078@fez.hyperreal.org>
In-Reply-To: <20050111155744.O7078@fez.hyperreal.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Here is the most recent exchange that I have had with Bruce Miller from 
NIST, one of the authors of the "public domain" JAMA package that we 
want to incorporate code from in Jakarta Commons Math.  I need a 
recommendation on how to proceed.  I shared excerpts from the posts to 
this thread with Bruce and it looks like NIST and MathWorks are both 
sympathetic, but unwilling to attach "apache friendly licenses" to code 
that they view as in the public domain. Personally I see no real risk in 
here, but this is not my call to make. Thanks in advance.

---------------------------------------------------------

Phil Steitz wrote:

 > Thanks for the quick response!  See interspersed.
 > Bruce Miller wrote:

[...]

 > Yes, as long as the copyright is essentially "given up" to public 
domain or a "friendly" license is attached.
 >
 >> (while Moler, of course, relied heavily on his past coding, 
including for Matlab...)
 >>
 >>
 >> I'll happily run the question by my boss, Ron Boivert (also involved 
with JAMA),
 >> but it would help if I understood the question myself, if not the 
answer :>


Well, I ran the issue by Ron, and by Cleve Moler (at MathWorks),
and others in the project.  Cleve doesn't want to add/change licenses
because he'd have to get MathWorks lawyers involved, who are already
swamped with other IP stuff...  All round, there seemed to be a "guess
they'll just have to trust us" sentiment.

As a linux (&, apache, other OSS) user, following SCO and other
developments, I am perhaps more sympathetic.  But, in the end, I guess
I'm not that clear on where the holes in "public domain" are, and how
to clarify them here.

Or in general: we (at nist) are generally supposed to release our
work (assume solely NIST work for purpose of argument) as public
domain --- our work is not subject to copyright.  IANAL, but
my interpretation is that I then cannot even add any license (GPL or
otherwise) to that software since I would seem to have to assert
"ownership" before I can "grant" any rights.... Or?

Anyway, back to the subject at hand;  Can we just interpret "public domain"
to mean what we usually think it means?

-- 
-- 
bruce.miller@nist.gov
http://math.nist.gov/~BMiller/

---------------------------------------------------------------------
DISCLAIMER: Discussions on this list are informational and educational
only, are not privileged and do not constitute legal advice.
---------------------------------------------------------------------
To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
For additional commands, e-mail: legal-discuss-help@apache.org