MR. LO;:  Good morning, Mr. Commissioner.  My name is Allen
Lo.  I'm also another example of a casualty from Group 2300. 
I worked as an examiner in Group 2300 for about two and a half
years, examining patent applications involving computer
control systems, computer-aided product manufacturing, and
error correction and detection systems.
Last March I left the PTO to work for the law firm of
Finnegan, Henderson, Farabow, Garrett & Dunner, where I
currently prosecute and write patent applications, much of
which involves software-related inventions.  I currently
attend the Georgetown University Law Center as a third-year
evening student.
Today I'm speaking on my own behalf.  The views that I express
today are my own and not the views of the firm or its clients.
Mr. Commissioner, I'd like to address two different matters
this morning.  First I'd like to speak about the group's
policy regarding the patentability of claims drawn to software
stored on a disk.  Second, I'd like to supplement the comments
that Dave Clark made about what the PTO can do to improve the
quality of examination based on my own experience.
Beginning with the first issue, it's been the policy of Group
2300 that claims drawn to software stored on a disk are per se
unpatentable.  During the examination of an application
involving a software related invention, examiners in Group
2300 place claims into one of two groups:  implemented and
nonimplemented computer software.
Implemented computer software generally refers to computer
software that's claimed as being executed on a computer, which
Group 2300 treats as being patentable, subject, of course, to
the novelty and nonobviousness requirements.  For example, a
claim reciting a general purpose computer running a novel and
nonobvious computer program is treated as being a new machine,
and thus would be allowed by Group 2300.
In contrast, nonimplemented computer software refers to
computer software that is not executed on a computer.  In
other words, simply a static program.
Claims reciting nonimplemented computer software may be
directed either to the computer program itself, such as a
computer program comprising followed by either source code or
means plus function language, or to software that's stored on
a disk, such as a computer-readable medium, storing a computer
program comprising, followed by source code or means plus
function language.
Group 2300 views these claims, the nonimplemented computer
software, as per se unpatentable.
I'd like to focus my comments on one particular type of
nonimplemented computer software, and that is claims reciting
a disk that store a computer program claimed in terms of means
for performing a function, say a function -- means for
performing function A, means for performing function B, et
Examiners in Group 2300 are trained to reject this type of
claim under 35 U.S.C. Sections 101, 102, 103, 112 first
paragraph, and 112 second paragraph.  These rejections can be
simplified, and I would generally classify them into three
different categories.
First, these kind of claims are rejected under 35 U.S.C.
Section 101 as being directed to printed matter.  The second
category, the claim is rejected under 35 U.S.C. Sections 101,
102, and 103 over a prior art disk by effectively reading out
any specific recitations in the claims directed to the
computer program, and then concluding that either the claims
are anticipated by a prior art disk, or that storing any type
of data on a disk would have been obvious.
The third category of rejections are under 35 U.S.C. Section
112 first and second paragraphs, because the disk itself is
unable to perform the recited functions, but requires a
computer to actually perform the functions, and therefore is
either indefinite, or the specification doesn't disclose how
a disk can perform the functions.
I believe that these type of rejections are either
unsupportable under the case law, or can be easily drafted to
avoid these kind of rejections.
First, with regard to the printed matter rejections, these
type of claims do not attempt to claim the mere arrangement of
words, which is really what the printed matter rejection is
all about, such as the program code itself.  But rather, these
type of claims specifically are directed to the functionality
that is provided by the computer program.  And thus the claim
really doesn't contain printed matter.
However, even if the computer program could be analogized to
printed matter, the case law does provide an exception to the
printed matter rule, which is that if there is a functional
relationship between the printed matter and the medium that
the printed matter is stored on, then those claims are not
considered printed matter -- printed matter rejections are not
applicable to those types of claims.
In the case of a computer program stored on a disk, the
computer program really transforms the disk into new disk kind
of the same way that a computer program transforms a general
purpose computer into a new machine.
With respect to the rejections over a prior art disk, it's
simply improper for the PTO to ignore any limitations in the
claim, particularly in this case where the computer program is
claimed in means plus function language.
And finally, with respect to rejections based on the disk
being unable to provide the claimed functions, the claim can
be drafted to be more specifically and particularly claimed if
actual function performed by the disk.  For example, the claim
could be drafted differently, and rather than being claimed as
a disk storing a computer program comprising means for
performing a function, means for performing a function, it
could be claimed as a disk storing a computer program
comprising a means for instructing a processor to perform the
function, means for instructing a processor to perform another
So in this way it actually is claiming what it actually does,
which is really to instruct a processor.  And I think that can
avoid those types of rejections.
It should be noted that the claiming of software on a disk is
not simply a trivial exercise in claim drafting.  Patentees
have an interest in obtaining claims drawn to software stored
on a disk.
By disallowing these types of claims, patentees must obtain
patents with claims drawn to software that is run on a
computer.  When patentee seeks to enforce this type of patent,
manufacturers and sellers of infringing software would not be
liable for direct infringement, but rather it would be the
users of the software that would be liable for direct
infringement by virtue of the fact that they're running the
software, because that's what the claims really recite.
Manufacturers and sellers of infringing software would not be
liable for direct infringement, but instead they would only be
liable to the patentee under some cumbersome theory of
contributory infringement, or inducement infringement,
requiring the patentee to prove additional elements, including
knowledge and intent.
Whether or not the PTO changes its policy toward
nonimplemented computer software, I think it's important that
the PTO at least publish in the official gazette clear
guidelines and rules defining the types of software claims
which they PTO considers to be acceptable.
I believe that a lot of these guidelines are not published. 
And so a lot of this information I have is only as having been
an examiner in the group.
Turning to the matter of improving the quality of examination,
I'd like to make the following observations and suggestions.
First, as others have suggested, examiners in Group 2300 need
to receive more legal and technical training.  Many of the
examiners who attend law school eventually end up leaving the
PTO to accept more lucrative positions in private practice,
resulting in fewer and fewer examiners with legal training in
the PTO.
I would recommend at least more in-house legal courses that
teach basic legal skills, such as legal research and writing,
be offered to those examiners who don't attend law school.
Further, examiners should be invited and encouraged to attend
meetings and lectures that relate to software patenting, for
example, such as today's hearings.  Yesterday I attended, and
I didn't see any patent examiners.  I don't believe that they
were actually notified of the hearings.  And I think
attendance at these kinds of things would be helpful to
examiners, at least so that they can understand the big
picture and appreciate the issues that they are actually
facing during examination.
Further, more technical courses need to be offered to increase
the level of technical understanding within the group.  It is
awfully difficult for an examiner to appreciate the advantages
of an invention when they don't really understand what it is.
And as Dave Clark pointed out earlier, the training of
examiners is undermined if the PTO is unable to retain its
examiners.  My experience has been that examiners leave the
PTO for various reasons.  Certainly many examiners are lured
from the PTO by the higher salaries that are offered by patent
law firms.  This problem could be alleviated somewhat by
raising salaries in groups with high turnover rates, such as
Group 2300.
However, I believe that some examiners, myself included, leave
the PTO because they feel the examining function is no longer
stimulating or challenging.  Finding solutions for retaining
these types of people may be difficult, but I believe that,
for example, the Examiner Enrichment Program which you
mentioned yesterday is a definite step in the right direction.
Finally, there is natural tension between the count system and
having a high quality of examination, as you mentioned.  I
think the count system is something that is necessary.  I
think you had mentioned reevaluating the number of hours,
perhaps, that the examiners should spend on a particular case. 
I think that would be helpful.
Some of the suggestions that Dave made earlier about having
group meetings and that kind of thing, and how that doesn't
really count towards an examiner's performance, can -- those
types of meetings can be counted by the fact that PTO
oftentimes does offer write-off time.  So time that people
spend in these types of meetings, they're not really held
accountable to produce additional cases.
I'd like to thank you for allowing me this opportunity to
testify, and I can answer any questions you may have.
COMMISSIONER LEHMAN;:  Thank you very much.  I don't think I
have a need at this point, but I think that was very helpful. 
And it's very helpful, the perspective of people like you who
have been in the corps, who are young attorneys or
attorneys-to-be, because you're exactly the kind of people we
need to know what's going on with.  Thank you very much.
Our next witness is Professor Samual Oddi of the Northern
Illinois University College of Law.
Back to the index of speakers for Arlington

Forward to Samual Oddi