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

[Joseph Kiniry]

Hi Niall,

--On 20 March, 2005 13:50:12 +0000 Niall O Broin <niall at linux.ie> wrote:

> On 20 Mar 2005, at 12:30, Joseph Kiniry wrote:
>
>>> No.  Because so many patents are held on what would otherwise be
>>> termed
>>> basic techniques (eg: the XOR trick to highlight an area of the
>>> screen),
>>> it is extremely difficult to write unencumbered software.  The trouble
>>> with the current system is that patents are held on many methods and
>>> devices which are trivial yet essential.
>>
>> How does this contradict my point?  Even the technologies covered by
>> "basic technique" patents are easily implemented in other, innovative
>> ways.
>
> The problem with "basic technique" patents is that they ARE so basic and
> are often NOT easily implemented in other, innovative ways.

In my experience I have yet to discover a "basic technique" patent that was 
not easily implemented in other ways.  Perhaps you have an example?

> The XOR trick
> is a very good case in point (though that patent was granted in 1978, so
> I presume it's expired by now) because it is so basic, obvious and simple
> to implement, and importantly when it started being used, very cheap in
> CPU cycles. I don't know if sprite based systems infringe that patent,
> because I think many of those use a hardware overlay plane. I do know
> that Autodesk was a licensee of that patent for AutoCAD, and I imagine
> other CAD vendors and other large visible targets using this idea were
> too.

The XOR patent was never actually contested in court to my recollection. 
Had it been I firmly believe that it would have been rejected given I saw 
work from the early 60s that discussed the technique (e.g., Sutherland's 
thesis).

>> Note that I *do* think that the existing US patent system is extremely
>> broken, particularly with regards to software and business patents.
>
> Indeed - one click comes to mind straight away. I grew up as an engineer
> being told that a patent was on the implementation of an idea, not on an
> idea, but those halcyon days are gone.

Agreed.  But even though the system is seriously broken, that should not 
prevent me from filing a patent that I believe is of high quality, fully 
implemented, properly researched, and (very) non-obvious.  I and my 
companies are not hypocritical: no matter how screwed up the system is I 
will not "take advantage" of it at the cost of my personal morals.  (Note 
that the same holds true from my POV wrt taxes, social services, etc.)

>> History has shown that existing trivial patents are voided once they
>> reach the courts or once an enormous corporation forces the USPTO into
>> a review. Of course, this is often a quite costly process.
>
> Exactly. And in the U.S., where in the majority of cases AIUI each party
> pays its own court costs, simply being threatened with a court case will
> make most companies cave in, because even if you win the court case, you
> lose financially. I wonder what's the general practice in patent
> infringement cases in Ireland, where in general in civil cases the losing
> party pays the costs of both parties.

I do not know what the general practice is here; all of my experience is in 
Holland and the USA.  It is an interesting question.

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