[OpenISO] Patents, OpenISO, and RAND licensing

[Norbert Bollow]

David Sugar <dyfet at gnutelephony.org> wrote:

> No...
> 
> Let me rephrase that...
> 
> HELL NO!

David, it's good to e-meet you again!  I agree wholeheartedly with
your sentiments here.

> Okay, I will elaborate...
> 
> When we speak of international standards, especially those that may 
> become mandated by law or otherwise required to enter and participate in 
> markets, these are fundamentally incompatible with patenting, even under 
> RAND.

Yes.  Whenever there is legitimate reason to consider something for
standardization which has been patented, that means that a situation
has occured where the patent system has led to an absurd result.

Unfortunately, although we have the right to create a standardization
organization, we lack the authority to fix the bugs in the patent
system which sometimes lead to such absurd results.

If (in what the ISO/IEC directives call "exceptional situations")
really no technically reasonable way around the patents exists, it
can turn out that the choice is between accepting a standard with
"RAND" licensed patents or not having any standard at all.  I do
believe that enough non-patented programming ideas exist already
that any decently skilled programmer's ability to solve any
programming problem by means of non-patented ideas is limited only
by the practical problem that (because of far too many patents,
unclear patent claims etc) it is impossible to know precisely what
is patented.  Therefore I believe that if ISO/IEC was applying
their directives correctly, no patent-ecumbered "standards" should
get accepted as ISO/IEC standards anymore.  By contrast, in the
context of hardware-implementable encodings, I know an example
where  the only (known) reasonable way to do the ecoding is patented.
In that kind of context, the ISO/IEC directives "exceptional
situations" rule is the best possible course of action to get a
usable standard without waiting until the patent system is fixed or
the offending patents have expired.

For information technology ideas which are typically implemented in
software I think it is clear that we must insist that nothing is
acceptable as a "standard" which is in any way patent encumbered.
(Patents are IMO not a problem if they come with a LGPL-compatible
free patent license or with a patent non-assertion covenant that turns
them into purely defensive patents.)

Open questions for OpenISO.org:

1) Should OpenISO.org require all patent holders to execute a patent
   non-assertion covenant, or should OpenISO.org offer the possibility
   of issuing a LGPL-compatible free patent license as an alternative?

2) Should OpenISO.org require all participants in the standards
   development and/or evaluation process to grant a patent
   non-assertion covenant (or, if that is considered an acceptable
   alternative, issue a LGPL-compatible free patent license)?

Greetings,
Norbert.


-- 
Norbert Bollow <nb at bollow.ch>                      http://Norbert.ch
President of the Swiss Internet User Group SIUG    http://SIUG.ch
Working on establishing a non-corrupt and
truly /open/ international standards organization  http://OpenISO.org