[ILUG] Re: ILUG sends s/w patents briefing document to Irish MEPs

[Joseph Kiniry]

Hi Paul,

--On 20 March, 2005 13:30:55 +0000 Paul O'Malley <ompaul at eircom.net> wrote:

> Joseph Kiniry wrote:
> [big snip]
>
>> primary factor in giving evidence to void a patent on the basis of
>> "obviousness" is the frequency of its use in existing software
>> systems.  E.g., one would guess that every sprite-based video games of
>> the past 20 years uses the XOR method previously mentioned.  This
>> means that FOSS developers, IMO, need not be aware of each and every
>> "obvious" patent because, unintentionally, their violation of said
>> patents *supports* the case against the patent in the first place.
>>
>> Joe
>
> Act 1 Scene 3
> [Scene opens to Paul getting into flame proof suit]
>
> Hi Joe,
>
> Perhaps you would be generous enough to allow me claim that it is
> innovative to extrapolate from the last couple of lines, that you, have
> proven, without intention that, the case for any patent is flawed. After
> all once solved any problem's solution is obvious. By extension given a
> smart enough person and the right tools the most efficient answer will be
> found. All that is really going on when problems are solved or new
> situations created is the marrying of not obvious partners, and that, I
> am happy to say is very obvious.

I strongly disagree with this statement.  There even exists work that I 
have done myself that is non-obvious (or even moderately difficult!) after 
I have completed it.

I use what I call the "Game Show Criteria" to evaluate a patent: You give 
me a description of the problem that need be solved in a problem domain in 
which I am a domain expert.  I then get 10 minutes to explain every 
possible generic solution that I can think of.  If I did not come up with 
something that has moderate overlap with the claims of the patent, then it 
is probably non-obvious.

Of course, my 10 minutes is usually 2-4 years for a USPTO reviewer...

> However I digress. If you do patent software, at what cost can one create
> such a patent, or on the other hand defend such a patent.

One can obtain a patent for what I consider a "reasonable" sum give the 
bureaucracy involved.  In my experience the intellectual effort and time 
involved are worth far, far more than the patent application costs itself.

> Then there is
> the case, when you come up against the large company with thousands of
> patents you must cross licence, and have your 'protected' work taken from
> you to be used by others, what good the costs you have incurred now? At
> this point the question of how do you defend your innovation must come
> into focus sharply.

Are you a Philosophy major?  :)  I have to re-read your sentences a few 
times to get their gist.

I do not cross-license my patents.  I have either exclusively licensed them 
or plan on making them freely available to everyone.

> From the above it seems we can reasonably suggest that software patents
> block progress after all that is what innovation is. Perhaps patents do
> it better than the supreme soviet ever could have wished to in the days
> of USSR.

Do you believe that patents block progress in fields other than software?

Thanks for these excellent discussions everyone!

Joe
-- 
Joseph R. Kiniry
Dept. of Computer Science, University College Dublin
http://secure.ucd.ie/