One might ponder as to whether or not
software performing the
encoding and decoding of images at
issue in the Microsoft v. Motorola case
discussed in the
previous Groklaw article on Judge Robart's ruling are
abstract
ideas in the above sense.
It seems that most software
patents we encounter here on
Groklaw do not represent "invention" at all, but
rather, are
patents on the ordinary work product of software engineers.
For
me, as a software engineer, that is the aspect I find
most outrageous.
However, if there is an area of software that could be
called
"invention", it must be in the area of data
compression. Many incredibly
wonderful schemes have been
invented over the years, beginning with
Lempel–Ziv–Welch
(LZW) and continuing on to the most exquisite algorithms for
encoding 3D video.
Then again, these algorithms come the closest to
being
pure mathematics as well, and in that way are no more
exquisite or
inventive than papers the are published all the
time in the area of
Mathematics, so I'll not be making an
argument for their patentability
here.
Instead, I would invite you to consider a different
aspect of
data compression, the nuts and bolts of coming up
with an implementation that
gives real time performance.
Back in the day when our PCs ran an Intel
286 16 bit
microprocessor running at 12 MHz, real time decoding and
playback
of compressed audio and video was a daunting
challenge, and real time encoding
was out of reach for many
things we take for granted today. The algorithms they
developed back in those days to get a workable level of
performance out of
codecs was a marvel to behold, and of
course written in hand-coded assembler
language.
Take the evolution of algorithms for efficient
implementation
of the FFT for example. They gave rise to a
new era of very useful
applications.
In the two examples given above, codec and FFT
implementations, I pay homage not to the algorithms, but
rather the
implementations as being truly inventive in terms
that might conceivably meet
the criterion to be patentable
and worthy of patent protection.
I
would be interested in your thoughts, and the thoughts
of others on that. (I
would add as well, that if there was
ever a case where source was needed in a
patent it would be
in such a case, where implementation is the invention, not
the algorithm.) [ Reply to This | Parent | # ]
| 
