I am starting to notice a growing number of people who claim that what
makes software Open Source is what *copyright* license agreement it is
licensed under.  This is not in fact the case: a program qualifies as Open
Source if the "distribution terms of open-source software must comply with
the following criteria" as described in the Open Source Definition:

  In practical terms this means that any exclusive rights that exist on
the work must be licensed according to this criteria.  While all the
licenses in the OSI approved license list deal with copyright issues, not
all of them address other exclusive rights like patents at all.

  I suggest clarification in the definition on this issue, including 
possibly a reference to a document that describes the issue in more 

 PCT - Patents, Copyrights, and Trademarks (and other Sui Generis

  I don't think Trademarks or other Sui Generis Protections are an issue
for Open Source at this time.  Requiring that someone rename the project
if they make a modification does not go against any of the criteria that I
can see.  I don't know how database protections are going to affect
collections of programs (distribution CDs, etc), but I can't think of a
problem that isn't dealt with by #9.  If I am wrong, please correct me --
are there discussion papers on various Sui Generis Protections already?

  Copyright is dealt with extensively on the site already, and is the core 
of the documentation.

  Patents do not seem to be dealt with generally, only specifically within
certain license agreements that mention all PCT issues in the single
license agreement.  A great example of a license that deals with the
patent issue is http://www.opensource.org/licenses/osl.php at "2) Grant of
Patent License".

  Not all OSI approved license agreements mention patents at all, leaving
a marketing loophole where someone can claim that software is Open Source
because it uses an OSI approved license, and yet is not open source
because it uses a patent license which is not OSI approved.  Unsuspecting
Open Source developers and users could then offer support for this
software only to find out after the fact that someone else claims
ownership (via patents) over their work.

  My reading of the OSD suggests that only an RF (royalty free) patent
license with no "Field of Use" restrictions would qualify.  The RF issue
is clear from #1, and this came out clearly in the discussions with the
W3C.  What I don't see clearly stated is how #3,#6,#7,#8 all add up to a
requirement that there be no "field of use"  restrictions, something that
wasn't clarified well during the W3C patent policy debates.  

  Licensing, including for derivative works, must be automatic (that any
recipient of the software receive the same patent license) in my reading
of the OSD #7.  This is one issue that many people incorrectly believe is
a GNU GPL issue rather than an issue with all Open Source compatible
patent licenses.
  (Example:  see the misinformation in
http://zdnet.com.com/2100-1104_2-5125160.html )


   I have been trying to deal with this issue in the context of IBM and my
work with Industry Canada on software patent issues
<http://www.flora.ca/patent2003/>.  We need to somehow encourage IBM to
come clean on how they appear to be adopting Open Source via using Open
Source qualifying copyright licenses while at the same time promoting
"unlimited patentability" worldwide <http://swpat.ffii.org/players/ibm/>
and not fully adopting Open Source patent licenses (such as the W3C and
now IETF and OASIS problems).

Here is an example "letter to the editor" I have sent in:
  IBM Leads in Patent (arms) Race

Note: There are all these Halloween documents discussing the OSI
battle-of-words with Microsoft, but I wonder why there is no similar
discussion with IBM?  Microsoft may be fun to play with, but they are
simply not the only organization that is committing actions that can harm
the Open Source movement.  I think IBM represents a considerably greater
threat than Microsoft, and IBM may turn out to be easier to reach to turn
them around.  If we can turn IBM around we may be able to turn around the
whole "information process patents" issue once and for all.

 Russell McOrmond, Internet Consultant: <http://www.flora.ca/> 
 Governance software that controls ICT, automates government policy, or
 electronically counts votes, shouldn't be bought any more than 
 politicians should be bought.  -- http://www.flora.ca/russell/

license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3

Reply via email to