🚀 go-pugleaf

In article <3fe8s8$56s@sparky.sterling.com> Adrian_Glanvill@stockley-park.sterling.com (Adrian Glanvill) writes:
>PMJI, but the fact is that some software patents _are_ ridiculous. The LZW
>patent has been mentioned here, but I believe one of the multimedia software
>vendors een patented the concept of "multimedia" - I think the definition went
>along the lines of a searchable database consisting of images, sound, animation
>and text. Correct me if I'm wrong but the patent was actually awarded to them.
>And that is undeniably ridiculous.

The Comptons patent was broad, but certainly didn't cover all multimedia
or the concept of multimedia.  But it's an example of how the patent
system can correct its mistakes without expensive litigation -- the
Commissioner of Patents, Bruce Lehman, ordered a reexamination of the
patent.  The reexamination will result in a limiting of the scope of
the patent, if not its cancelling.

(Others can request reexamination by submitting prior art in the form of
printed publications or issued patents, indicating how the art pertains
to the patentablity of the invention, and paying a $2,320 fee.  Unlike
in litigation, there is no presumption of patent validity in a reexam.
Congress considered legislation late last year to improve reexamination
by allow more participation by the requestor.  It's likely that it will
be considered again this Congress.)

I'm told that, despite all the fuss about the patent (caused by Comptons
bragging about it and the press not understanding what was really covered
by the patent), nobody submitted prior art to the patent office as a basis
for the reexamination.


>And last but not least, it's a good job whoever dreamed up the algorithms for
>long division and long multiplication didn't patent them, or every school in
>the world, and every person who ever did maths would be an infringement.

This is a red herring often raised as a criticism of patents.  Even if
caveman Og was the first to invent the procedures for long division and
multiplication, and had received a patent on them because they were novel
and nonobvious at the time of his invention, the patents would have expired
long ago.  He would have benefitted from his invention, if he could have
commercially exploited it, and society would have his teaching of the
novel and nonobvious techniques.

Even if Knuth had been able to patent every algorithm in his three volumes,
they would now be in the public domain because the patents would have
expired.  The "what if <person> had patented <old thing>" argument doesn't
seem to recognize that patents have a limited duration.

You may argue that the duration of a patent is too long when compared to
software life cycles, but that's a different argument than the nonsense
one given above.

				Lee A. Hollaar
				Professor of Computer Science
				Registered Patent Agent