[IxDA Discuss] Microsoft to license Office 2007 UI system

[Mark Bardsley]

Alan Wexelblat wrote:

> - patent examiners are not skilled in the art.  At best they're
> generally familiar with computers and software.
> - patent applicants are not required to disclose prior art.  OK,
> technically they should but ignorance is bliss.  If you fail to
> disclose then it's on the examiner to find prior art and if he does
> find some all that happens is you need to resubmit.  There's no
> penalty for - let's be frank - lying by omission.
> - software patents have PATHETIC levels of prior art reference.  If
> you doubt this, do just a few seconds of searching on the USPTO site.
> Go ahead, it's public on the web at
> http://patft.uspto.gov/netahtml/PTO/search-bool.html

I asked a friend's perspective who is a patent attorney that has worked at
Silicon Valley law firms representing large and small software clients and
is currently a corporate patent attorney at a large company that has been
mentioned as part of these discussions. 

Before I quote him he also wanted me to state that the views are not his
"personal views, are not legal advice, and do not necessarily reflect the
views or positions of my employer, past clients, or future clients."

So here are his thoughts:

"1.  Duty of patent applicant to disclose prior art that may be material to
patentablity.  Here, I think he misstates the law.  In particular the poster
says "patent applicants are not required to disclose prior art".  this is
incorrect.  Patent applicants (and their attorneys) are under a legal duty
to disclose all prior art that may be material to the patentablity of the
claimed invention that is not merely cumulative with the art already of
record.  In this regard, you might direct the poster to 37 CFR 1.56.  The
penalty for violating this duty is severe.  Specifically, if it is found
that the patent applicant intentionally didn't disclose prior art material
to patentablity for a patent application, the resultant patent is
unenforceable.  That is, basically all rights of the patent are gone if the
patent is procured by the applicant concealing material prior art.  I think
it is fair to say that in basically all patent litigation, the accused
infringers very, very thoroughly investigate the prior art the inventor knew
about and check to see that the inventor properly disclosed this.  Failing
to comply with this duty under rule 56 is called "inequitable conduct".
Invite the poster to search the internet, including case law from the court
of appeals for the federal circuit, regarding inequitable conduct."
 
"2.  Citation of prior art in software cases - Here, the poster's anaysis is
lacking.  The poster states "software patents have PATHETIC levels of prior
art reference" and "proves" his allegation by noting the number of
references cited in a single software patent.  I think the quality of a
prior art search is determined by whether the searcher found prior art that
would anticipate or render obvious the claimed invention.  That is, a
searcher could find 10,000 references (including research papers, books,
IEEE papers, foreign language documents, and the like) that mention TCP/IP,
but that aren't really directed to the specific problem solved by the
subject matter of the patent application or details of the solution of the
patent application.  Despite having a large number of references from many
sources, I wouldn't regard this as a particularly good search.  This type of
search could be done on google in about 10 seconds.  On the other hand, a
searcher could find 2 or 3 references that were published before the filing
date of the patent application that either disclose the patent application's
solution, problem addressed, or components of the solution.  This search
would be more helpful to determining the question of patentablity.  In
short, I think the true measure of the quality of a search is whether there
is better prior art out there than that cited by the examiner or disclosed
by the applicant."
 
"3.  Related to my point #2 above, for the prior art to be helpful in
determining patentabity, whether it was published a day before the invention
or 100 years before the invention, the law considers it the same.  So the
fact that age of the oldest reference cited is much less than the age of the
underlying technology shouldn't be a concern.  In fact, for inventions that
are really innovative, you would think that if they were known by others
that they would have only been known at most a few years before.  20 year
old prior art is typically not that helpful for software patents due to the
fast pace of innovation in this area."
 
"4.  I'm not saying the art cited in this particular Microsoft patent is the
closest prior art or that a competent search was performed.  I am saying
that this poster hasn't said anything to make me think that the search was
inadequate.  Invite the searcher to find a reference published before the
filing date of this patent that discloses the features recited in this
patent's claims in the same manner they are recited in the claim that was
not cited by the examiner.  If the searcher can do this, then they can argue
that the search wasn't adequate.  Without such a showing, I don't think they
are saying much."
 
"5. Having worked on software patents, it is my sense that the U.S. Patent
office's ability to search software patents has improved a lot over the past
8-10 years.  It's not perfect, but they do what they can given the time and
resources they are given."

- Mark Bardsley