MR. GIMLAN:  Honorable Commissioner, distinguished members of the panel, may
name is Gideon Gimlan, and I do not come here to represent any particular
organization.  It's true that one of the labels I wear, if you want to
define where I am coming from, is that I am a patent attorney with the law
firm of Fliesler, Dubb, Meyer & Lovejoy of San Francisco and Sunnyvale. This
particular firm represents numerous high technology companies located in
Silicon Valley and elsewhere.  The work of the firm and my own work includes
the preparation and prosecution of software_related patent applications in a
variety of areas, including networked computer systems, graphic imaging
systems and mainframe computers.

I have to add the immediate legal proviso that these comments are my own
personal views based on general experience, and not those of any member of
the law firm or of any clients represented by the firm.

I come before you wearing an additional label __ this is part of my general
experiences __ that prior to becoming an attorney, prior to so_called
defecting into law school, I was also an engineer who worked in the field
for over seven years.  I would characterize the nature of the work that I
did as being a hardware/software engineer.  And the reason I use that
characterization is that a lot of the work assignments that I followed
through with included the step of choosing whether to implement particular
functionalities in software or hardware.

Insofar as the experience I've had from that background, I'll have to repeat
what Ron Laurie so eloquently phrased, is that there is a spectrum,
continuous spectrum, in terms of what we define as hardware and software,
and it's almost impossible to cut that spectrum in half and define some line
that separates something from being hardware or software.

Also, while I'm on that topic, it brings back to mind while I was working as
a hardware/software engineer, Mr. Fiddler, who was here before, mentioned
something about word processing being a old and obvious technique that
shouldn't be patentable. I unfortunately go back to the days when people
were doing affordable word processing with hard_wired machines back in the
early '70s.  The original versions of affordable word processing came in the
form of the IBM magnetic card, and there were a lot of companies who came
out during that time and started to produce hard_wired word processing that
eventually led to software types of devices.  Generalized computer has taken
it over, but the origins of it really lie in hardware in terms of having
affordable word processing capabilities.

The question that I really wanted to focus on here today was Question No. 3
in your requests for comments: What are the implications of maintaining or
altering the current standards for patent eligibility for software_related

And I'd like to retitle that as "What is the current PTO practice?  And
where is it leading us to in the software arts?"

My own personal experience is that, insofar as anticipation and obviousness
are concerned, the examining corps treats software_related inventions no
differently than other kinds of inventions.  The legal tests for 102/103
determination are fairly well_established and most examiners treat
software_based cases with the same uniform fairness as hardware_based cases.

The issues of finding good prior art in software area is no different than
that in any other art.  As an aside, in terms of quality, I find that the
European patent office tends to find closer prior art for particular
inventions than does the United States Patent Office, but again, that
applies to general subject matter and is not specific to software_related

Insofar as Patent Office inquiries into 35 USC 101, what constitutes
statutory subject matter, I fail to see any across_the_office uniform
consensus on what is or is not statutory, the OG guidelines notwithstanding.

The treatment of statutory subject matter question appears to vary greatly
from examiner to examiner.  Some examiners are lenient in what they consider
to be statutory, while others seem to be on a witch_hunt for a 101 basis of
rejection.  This injects a considerable degree of uncertainty into the
application process.  You cannot predict the outcome of a 101 issue with any
degree of confidence.  It very much depends on which examiner you draw for
your case.

Perhaps "software_related" isn't the proper term for what I am trying to
address here.  The problem more properly fits under the broader rubric of
algorithm_related inventions and should the PTO be expending so much time
and energy trying to weed out claims that arguably extend or encroach into
nonstatutory areas.

I suggest that the answer is no.  The Patent Bar and Examining Corps are
wasting client money and taxpayer money arguing over metaphysical
abstractions.  That to technologists in the field sounds like we are
debating over how many angels dance on the head of a pin.  The case of In
re: Iwahashi serves as a good example.  It was not strictly speaking a
software_related case because the claim preamble started off with, "An
autocorrelation unit, dot dot dot, comprising."

But if one wished to take some license and rewrite the preamble to start
with, "A computer comprising," and I note that that was done in Example B of
the PTO request for comments, then in my mind this should not materially
alter the gist of the invention.

Any digital signal processor, including the one in Iwahashi, can be viewed
as a computing machine, or quote "computer" if you choose, one could then go
out on a limb to call each invention that uses a digital signal processor as
being software_related because its operations can be described in
algorithmic terms.

Notice that I didn't say controlled by a computer program or controlled by
quote "software".  There are those skilled in the art who will argue even
today that a computer program can be used as a description of the operations
to be carried out by the machine, and the description does not necessarily
have to form part of the machine that actually performs the described
operations.  The machine's control lines could just as easily be driven by
combinatorial logic as from a memory source.

In the end, it should make little difference that an invention is
implemented in hardware, software, or in_between_ware.  In the eyes of the
electronic circuits that carry out a given invention, there really isn't any
functional difference.  A set of electrical signals are first supplied to
the DSP machine.  Perhaps the input signals originate from a memory device
like a ROM or a floppy disk, perhaps they come from an x_ray machine. 
Irrespective of origin, the signals are somehow transformed by the machine. 
Then they are output, perhaps for return to memory, perhaps for routing to
some other immediate use, such as creating a real_time high_definition video

One inventor recently looked at me with bewildered eyes when I tried to
explain some of the 101 concerns related to his particular case, and he
said, "I don't understand, data is data, what does it matter whether it
comes from an x_ray machine or from memory?  What is government up to?"

And in that quote I've taken some literary license to replace what the
actual source of the question was, but okay.

I think the problem and the answer lie in how we as human beings come to
appreciate the subtle implications of a given invention.  We need to step
back and ask, Has the inventor come up with a faster or cheaper way of doing
things even if the improvement is found in software?  Has the inventor
compressed the physical size of an apparatus so that something smaller can
now do the job of something that previously had to be much larger?  Has the
inventor obtained a higher level of resolution than was previously feasible?

We see in hindsight that these kinds of improvements __ faster, smaller,
cheaper, better resolution __ have brought us the miracle of affordable
palm_top computers, ones that have pen_based graphical user interfaces, and
ones that, arguably, give even the technical neophyte access to the powers
of the digital revolution because of their intuitive nature.

I think we can all agree in hindsight that these are the kinds of
innovations that our patent system is supposed to protect and foster.  But
when we turn away from past glories and look to the next invention, we are
somehow daunted by the enigma of this thing we call software.  We are all,
in a sense, blind men beating at a pachydermial beast, each finding
something different based on the angle from which we approach it.  Some say
this software stuff is more like the punched paper in a player piano, or
like the music recorded on a vinyl record.  Others say it's more like the
mathematical proofs of their college calculus classes.  Yet others say it's
something that is still in its infancy, that will grow and evolve into
something we still do not fully understand.

Of course, in the meantime, software applications keep pouring into the
Patent Office.  So what should we do?  Should we tell those who craft new
software to go away?  You are not welcome at the Patent Office?  Should we
direct every algorithm_smith over to the line at the Board of Appeals?
Every examiner has his or her own personal angle on how to deal with this

This leads to a haphazard system which gives inventors __ particularly those
that have had the misfortune of being assigned to an "anti_algorithm" or
"anti_software" examiner __ the impression that they are not receiving
uniform, fair treatment.  It is absurd in the mind of many technology gurus
that an invention is okay if implemented in hardware but suddenly becomes
unaccepted because it is implemented in software.

The pat answer, of course, for inventors who face such examiners, is that
they can always go to the Board of Appeals, and if not satisfied with the
results there, they can go higher to the Federal Circuit.  But that doesn't
happen with regularity. What really happens is that many patent_worthy cases
fall by the wayside, not because the applicants agree with the examiner's
101 position __ and as a side comment, I sometimes wonder if even the
examiners themselves agree with their official position __ but because of
monetary considerations, it's just too expensive to go forward any further
and appeal.

One could argue that this problem could be taken care of by well_to_do
corporations, that they should lead the charge into the courthouse and help
us create better law, but that doesn't always work.  Some corporations are
afraid to get on the bad side of a key examiner.  Even those that are brazen
think twice about pouring more time and money into an application that is
already twice rejected by an examiner.  Most inventors, and corporate
executives for that matter, do not have the experience or patience to
grapple with the kind of metaphysical questions that are posed when a
Section 101 rejection is raised.  For example __

COMMISSIONER LEHMAN:  Mr. Gimlan, we're running out of time.

MR. GIMLAN:  Oh, I am, okay.  Then let me skip to my proposal then.  I think
that the ongoing witch_hunt at the patent office for nonstatutory subject
matter is in essence driving technology gurus away from the system.  They
simply don't understand it and will bypass the system.

My proposal is that unless particular claim in an application is clearly
limited to the practice of a mathematical algorithm, the Patent Office
should allow the applicant to disclaim within the body of the claim that
portion of the claimed system or process that falls outside the scope of

35 USC 101, and then allow the case to go to issue as is, assuming there are
no other bases for rejection.

After the patent issues, we should let experience and the advice of
technical gurus help us to decide whether an accused device falls within the
scope of a claim as interpreted under 101, or whether that accused device is
protected because in order to enforce the claim, you would have to transmute
its meaning such that it becomes a claim to a mathematical algorithm.

COMMISSIONER LEHMAN:  We can certainly take everything and read it over very
carefully with the specific suggestions.

MR. GIMLAN:  Okay, thank you, Commissioner.

COMMISSIONER LEHMAN:  We really thank you for sharing this with us today.

This idea that we have created a sort of artificial determination for patent
__ an artificial subject matter, in a sense, for patent lawyers to avoid
the, in order to deal with these 101 determination problems is clearly
something that we've heard from other witnesses here.  This is worthy of
looking into.

Next I'd like to call Tom Cronan, Secretary and General Counsel of Taligent,

Back to The San Jose Index

Forward to Tom Cronan