Newsgroups: comp.software-eng,misc.legal.computing,misc.int-property
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!pipex!uunet!world!srctran
From: srct...@world.std.com (Gregory Aharonian)
Subject: Review of testimony at Patent Office hearings in San Jose
Message-ID: <CKGvsu.1IM@world.std.com>
Organization: The World Public Access UNIX, Brookline, MA
Date: Sun, 30 Jan 1994 23:55:42 GMT
Lines: 430
REVIEW OF USPTO HEARINGS IN SAN JOSE
ON SOFTWARE PATENTING
Jan 29, 1994
Gregory Aharonian
Internet Patent New Service
P.O. Box 404, Belmont, MA 02178
617-489-3727 pate...@world.std.com
Last week the Patent and Trademark Office held two days of hearings at
the San Jose Convention Center, in which fifty testified for approximately
11 minutes each during the two days. Commissioner Lehman, along with about
a dozen other PTO officials, including Gerald Goldberg, director of Group
230 (which examines most of the software patents) were in attendance. About
125 people attended the first day, and 62 people the second day.
The testimony was quite interesting, which I will summarize briefly before
presenting summaries of each speaker's comments. I will also offer predictions
to what will happen to software patenting activities.
Summary
- With regards to software patents, most lawyers and large corporations were
in favor of allowing patents for software, while many programmers and small
companies were against awarding software patents, with interesting exceptions,
such as Oracle which testified against software patents. A few argued that
the growing overlap of hardware and software design tools made it increasingly
difficult to distinguish hardware and software from a patent claims point of
view.
- Everyone agreed that patent examiners and the patenting community need
access to a comprehensive software prior art database, to prevent awarding of
patents for non-novel and obvious ideas. Since I have the only large such
software prior art database, this constant refrain was music to my ears,
especially since I know how very complicated prior art databases are to
construct and maintain.
- Opinion was split on pre-publication, that is, publishing applications
some fixed time, say 18 months, after submission. In this way, the public
could offer evidence of prior art before a patent was issued. This is
standard practice for patenting in foreign countries.
- Most suggested that an open reexamination procedure would help greatly,
where prior art evidence submitted for a rexamination could also be used later
on in a court case, and that those requesting a reexamination be able to
participate in the proceedings.
- Most observed that patent examiners need more training, and that the
software engineering community should help out. Many suggested moving Group
230, the software/hardware examiners, physically to Silicon Valley, to
allow the Group to work more closely with its main customers.
- Many observed that current statutory regulations on how software and
algorithms can be patented are having little effect, due to inconsistent
application of the guidelines by patent examiners. Some suggested that these
guidelines be dropped completely, and allow any combination of hardware and
software to be patented.
- Many called for the Patent Office to make better use of the Internet for
distributing patent information and communicating with the public.
Much testimony was given to personal ancedotes about patenting software.
All of the testimony will be made available in February from the Patent Office
for sale in printed form for $30, or for free by anonymous-ftp. There are
more hearings in Washington in February.
Predictions.
As much as the League for Programming Freedom hopes, and argued somewhat
forcefully, there is no way patents for software will be banned. Too many
big software vendors desire software patents, including Microsoft (for which
the January 24th issue of Computer Reseller News has a two page article on
Microsoft's glowing clout in Washington). Beyond politics, it will be
impossible to legally define what is 'software' clearly enough to know what
to ban. Software patents are here to stay as long as patents are issued.
Shortening the life span of patents (down to as short as three years) will
also not happen, again because of politics and the impossibility of clearly
defining software.
Pre-publication will also not come to be, for the same reasons (whatever
they are) that first-to-file will never come to be (at least over the next
few years). Especially in the field of software, the Patent Office would not
be able to deal with the deluge of material sent in as challenging data, short
of getting lots more money from Congress or raising fees (though this also
plagues open reexamination).
Some statutory rules will be changed, because there are some obvious fixes
to improve practices. Personally I would like to see the guidelines in the
September 5th 1989 OG scrapped because they are not having much of an effect,
nor would adding more guidelines have much of an effect. Since these rules
changes don't require Congressional approval, it is a place where the Patent
Office can do something quickly.
The reexamination system will probably be opened up, with more public
participation and the ability to reuse materials submitted for consideration
in later court proceedings. I think this will cause a document overload
(until the prior art database is built), especially when public contributions
are solicited, as well as a massive copyright problem.
Efforts will be started to build a software prior art database. Though
based on comments heard at the hearings, few understand how complex a problem
this really is. This effort will be coupled somehow into public rexamination,
though will be greatly more complicated if the hardware/software equivalence
argument gains much acceptance or is used in a court case. And copyrights
of prior art documents will be a major problem for a government agency to
deal with.
In short, while there are problems with the patenting process vis-a-vis
software, things aren't too bad yet and there is time to effect changes to
manage the process better. A few years from now, without changes, there will
be a mess.
==============================================================================
SUMMARY COMMENTS OF TESTIFIERS AT THE HEARINGS
Wednesday, January 26
(in order of testimony)
Joe Clark, Chairman and CEO, Video Discovery
Frustrated by patent system. Afraid of pending patents kept secret until
issuance and the potential liability due to infringement. Wants to remove
secrecy of patents during prosecutions. Wants reexamination more public.
Wants hiatus on issuance of multimedia patents until these issues are
settled.
Ryal Poppa, Chairman and CEO, StorageTek
They have 300 programmers, and support software patents. Want decompilation
rights for interoperability. Oppose European efforts to use patents and
copyrights to limit competition. Wants law to allow access to source code
of APIs.
William Ryan, Intellectual Property Owners/ATT
Supports software patents. Copyrights useful for software expression. Courts
have made it quite clear that patents are for protecting software. ATT has
had good software patenting experiences. Patents useful for alliance building.
Software patents should not be treated differently. Against new statutes for
handling software. Current patent and copyright statutes sufficient and work
well. Better prior art information needed.
Richard LeFaivre, VP at Apple/ CBEMA
No need to treat software patents differently. As long as new and non-obvious,
it should be patented no matter how implemented. Standards for software
patents should be same as other technologies, not restricted. To restrict
would hurt US companies' protection overseas. Better software prior art
needed. Don't cut back on software patents.
Tom Lopez, President, Interactive Multimedia Association
Patent system casting cloud over industry. Powerful patents threatening in
integrated industries and customer application world. Don't like broad
claiming process/system level patents. Support patent system for software,
but should work fair and minimize speculation. Worried about weight of
patent examiner's review when trying to overturn.
Paul Heckel, President, Hyperracks
Very strongly support software patents. Helps protect small companies,
stimulates new technology introduction. Press is exaggerating patent problems.
Told a few stories of how small companies were able to raise money because
investors saw or demanded software patents.
Robert Kohn, General Counsel and VP, Borland
Argued that there is public confusion over issue of enforcement versus scope.
One Patent Office position implied more enforcement requires more broader
scope, especially for copyright protection. His comments reflected Borland's
loss to Lotus (under appeal) over copyright of 1-2-3 features.
Douglas Brotz, Principal Scientist, Adobe
Software should not be protected by patents. It is wrong to do so. Software
needs innovation. Software patent suits have hurt Adobe, which has spent
$4.5 million over five years successfully fighting off an infringement suit.
Industry leaders like Microsoft didn't rely on patents for their success.
Copyright confers appropriate levels of software protection. All software
is math algorithms in a reductionist sense which supposedly isn't patentable.
Delusion to think that patents protect small entrepreneurs. With regards to
interoperability, those that don't allow for it usually are losers, so Adobe
opposes any law to force decompilation rights.
Hans Troesch, Partner, Fish and Richardson
Wants qualified examiners and comprehensive prior art database. Lawyers can
get software patents, though painfully. Software patents should be allowed,
but only non-obvious and novel ideas. Should not have to play tricks with
claim writing to get a software patent. Copyright not sufficient for
software protection.
Robert Sabath, World Intellectual Property and Trade Forum
More flexibility needed for software patenting process. Don't want to slow
process through statutory rules. PTO maintenance fees too high. West Coast
deserves its own public search facility (similar in breadth to Crystal City),
and more throughout the country in schools and libraries. Need a comprehensive
software prior art database, one that is indexed. Relaxing rules for software
patenting and algorithms results in more software patent, but also more
software prior art to limit future software patents.
William Benman, Partner, Benman & Collins
Wants better software patent application examination, which is adequate now.
Software prior art problems no different that other technology areas. He
suggested a higher filing fee for quicker processing of applications.
Jerry Baker, VP, Oracle Corporation
Patents inappropriate for software. Copyright and trademark sufficient.
Software fundamentally different from other technologies, product cycles less
than patent processing times. Software rarely has quantum breakthroughs that
are worthy of patent protection. Patent protection should be eliminated for
software. Prior art database needed. Patent life for software should be
reduced to three years. Patent review should be more competitive. Software
patent examiners must be trained and paid well.
Carl Silverman, Chief Counsel, Intel Corporation
Software related technology should be patentable. Support current statutory
laws regarding software patents. Mistake to treat software patenting
differently. Can live with software patent pre-publication.
Kaye Caldwell, President, Software Entrepreneurs Forum
Members of her organization (about 1000 software developers in Silicon Valley)
are split on the issue of software patenting. Members feel PTO favors big
companies. Want infringement more clearly handled. Patents threaten small
companies. Improve the prior art database. Better train examiners on how
to deal with software patents, and increase their pay. Accelerated
examination should be improved. Prepare educational materials for press and
public. Use the Internet to help with prior art.
James Chiddix, VP, Time Warner Cable
Too many obvious ideas are being patented. Time Warner is the largest
copyright owner in the world. More rigorous enforcement of current novelty
and obviousness tests will help greatly.
Dennis Fernandez, Attorney, Fenwick and West
Two to three year backlog of patent applications hurts companies. Software
patents in the EDA industry do no significantly impact business. Companies
in the EDA industry are sophisticated, can design around other patents or
obtain reasonable licensing fees.
Pete Antoniak, Solar Systems Software
In the software industry for the "big" payoff. He was encouraged to apply
for a patent. Got a mechanical examiner for an educational software patent
application. His application was rejected three times, the first two for
reasons the original examiner admitted were unfounded.
Prof. Lee Hollaar, University of Utah
Software does not evolve quickly - Knuth's books, still used, are twenty years
old. Better software prior art database needed. Examiners are getting better.
Software industry not the best researchers in general, so lax about prior art
searching. Past PTO position of shunning software patents prevented building
a prior art foundation to reject future software patents. Open files during
prosecution after fixed period, after first office action. Section 101 battles
not worth fighting, better to shift focus to 102 and 103.
Steven Henry, Attorney, Wolf Greenfield & Sacks
Software industry no different than others. Typically, reexamination favors
patent examiner's position at the expense of challengers. Examiners need
increased training. People loosing faith in the system. Exceptions like
Compton getting the attention. Involved with company that couldn't get
funding without having filed apatent. Investors ask him about patent
protection often in regards to investments.
Sal Cassamassima, General Counsel, Exxon Production & Research
Want PTO to liberalize guidelines on math algorithms, of great importance to
Exxon. Seismic algorithms lead to large economic returns after large
investments in research and development. Dispense with Freeman-Walker-Abel
two part test.
Robert May, Iconic Interactive
Supports software and interface patents. Clients want indemnification for
prior art infringement, but search is expensive. Make more patent information
available over the Internet.
Neil Brown, Software Contractor
Against software algorithm patents, along with his colleagues. More software
patents will hurt innovation. Access to prior art is difficult, and prior
art determination could be intractable. Software patents are a blight.
Copyright and trade secrets sufficient.
Gordon Irlam, League for Programming Freedom
Prefer traditional "literal aspects" copyright doctrine. Oppose copyright
"look & feel" and "software patents". Big successful companies didn't rely
on software patents to grow in the software industry, like Microsoft.
Software patenting needs economic analysis.
E. Robert Yoches, Attorney, Finnegan Henderson Farabow Garrett & Dunner
All technologies have unique aspects, and all share much in common, including
software. Software patenting process can be improved with ideas from hearing.
Investors prefer patents over trade secrets, which investors feel can easily
become public. Software is easy to design around, so it does promote
innovation without restraining trade. Section 101 is an arbitrary handicap.
Jim Shay, Attorney, Morrison and Foerster
Software industry benefits from patents. Patent examiners don't have best
access to prior art. PTO should work with industry groups. Pre-publication
and open reexamination would help.
Thursday, January 27
(in order of testimony)
Jerry Fiddler, CEO, Wind River Systems
Stop issuing software patents. Copyright, trade secrets and trademarks is
sufficient protection for software. Software patents are harmful to country,
impedes innovation. Patents interfere with openness.
Jim Warren, Director, Autodesk
Software mimics the mind, and differs from all other devices. Extending
patents to software allows monopolies of intellectual processes. No
evidence that patents promote software progress. Autodesk is having to waste
money filing defensive software patents, and spends millions fending off
frivolous lawsuits. Intellectual/algorithmic inventions should not be
patentable. Freeze current applications. Reduce protection period.
First-to-file is good. Mandate cross-licensing for two years.
Michael Glenn, State Bar California
Need more qualified examiners. Need better software prior art databases.
Open up Internet to the APS text databases. Respect non-obviousness and
novelty of new software ideas.
Paul Lippe, General Counsel, Synopsys
First to market is a high advantage equal to patent monopoly protection.
Don't need strengthened patent laws for software. Most software patents are
incremental. No good EDA technologists work for Patent Office. A growing
problem for patent world as hardware/software converge.
Tim Boyle, Director, Multimedia Development Group
Wish for some sort of peer and academic review involvement with software
patent applications. Want a better software prior art database.
Ronald Laurie, Attorney, Weil Gotshal & Munges
"Software patents are bad" is legally unsound. Being mathematical or logical
operations should be irrelevant to patenting. No distinction between hardware
and software, such as neural nets. Novelty and non-obviousness must be applied
rigorously.
Lee Patch, General Counsel, Sun Microsystems
Spend lots of time dealing with nuisance patents. System is broken, creating
too much uncertainty, affecting large financial decisions. Low quality
software patents being issued, due to prior art and non-obvious threshold
being too low. Court of Appeals has lower standard. Examination must be
improved. Life of patent should be from filing date to prevent secrecy of
patents causing problems.
Christopher Byrne, General Counsel, StorageTek
It is a mistake to deny patents to software, thou process has to be improved.
Improve the training for examiners. Prevent "submarine" patents delaying
processing through secrecy. Software patenting process works, but should be
improved.
Gideon Gimlan, Attorney, Fliesler Dubb Meyer & Lovejoy
No clear distinction between hardware and software. EPO generally finds more
applicable prior art.
Tom Cronan, General Counsel, Taligent
Patent stimulate investments in software. Hire more computer science graduates
to be examiners. Build a human software prior art database, using the
Internet. Wants pre-publication after a fixed time period.
William Neukom, VP, Microsoft
Current laws appropriate for software patents, but system can be improved.
Support improving reexamination process. Better software prior art database
is needed. More prompt processing of applications needed. Microsoft has been
target of infringement suits. Had no patent lawyers two years ago, will have
six by the end of this year.
(And frankly, isn't it scary for Microsoft to have a lawyer named Nukem :-)
Charley Morgan, VP, Prudential Insurance
Being threatened with patent infringement lawsuits over life insurance
software systems. Better prior art is needed.
Richard Stallman, League for Programming Freedom
Software patents impede software progress. Software evolutionary, no
quantum jumps, little novelty. Software R&D has different cost structure.
How many patents does GNU-C infringe? Probably way too many. PTO mistakes
causing gridlock. The country has had much progress without software patents,
why start now?
Timothy Casey, Patent Counsel, Silicon Graphics
SGI founded on a Stanford patent. Patents not a bad idea. Too many broad
software patents. Need more examiners with software backgrounds. Revamp
reexamination, and make process less expensive. Do pre-publication. Give
examiners better software prior art research tools. Pay examiners more.
Regionalize Patent and Trademark Office, moving Group 230 to Silicon Valley.
Robert Greene Sterne, Attorney, Sterne Kessler Goldstein & Fox
Software patents are proper. Patent lawyers profit from current uncertainty.
Investors want patent protection before investing. No lines drawable for
what is an algorithm. Everything is mathematical algorithms - look at CAD
systems. Functionality is important - not implementation. Claiming should
be flexible. Patent applications must be processed more quickly.
Victor Siber, Senior Counsel, IBM
3.5 percent of IBM software programs protected by patents. Helps with
foreign market access. Patent office should do novelty checks better.
Standards should not be altered for software patents. To do so will negatively
impact investment. Hardware and software too interrelated. "Sui generis"
rights would be devastating.
Micheal Lachuck, Attorney, Poms Smith Lande & Rose
Clever patent drafters will always get around any set of rules. Against
pre-publication. Reexamination procedures should be improved. Threat of
injunctive relief for liability due to infringement makes risk very
uncertain.
Greg Aharonian, Source Translation & Optimization
Brilliant, riveting testimony that kept them spellbound for hours :-))))))
Seriously, I talked about the equivalence of hardware and software, and my
software prior art database. I estimate at least $10 million would be
needed to build the foundation of a comprehensive prior art database and
process for maintaining and upgrading such a system. Open reexamination,
while a good idea, will be a paperwork nightmare for the PTO.
Gregory Cole
Need more knowledgable patent examiners. Need software patents. Bring in
outsiders to help with examination.
Barry Graham, Int. Fed. Industrial Property Attorneys
Software patents help to acquire investment. Help protect commercial
successes.
Edward Lemon, Software Engineer, Network Computing Devices
Do not need software patents. Too many bad patents.
Roger Schlafly, Software Engineer, Real Software
In favor of software patents.
Russell Brand, Computer Science, Reasoning Systems
Against software patents. Lawyers and programmers expert at getting around
rules. "Look and feel" copyright protection also a problem.
Willis Higgins, Attorney, Cooley Godward Castro Huddleson & Tatum
Copyright juries have to act like patent examiners without the training,
so using copyright protection for software also risky and uncertain.