IP: A review of the EData patent situation

[Dave Farber <farber () central cis upenn edu>]

Date: Wed, 19 Jun 1996 10:15:40 -0400
From: srctran () world std com (Gregory Aharonian)




!19960529  How much bad software patents will cost you


    As many of you know, the first of the "threaten/sue everyone on the
Internet using a questionable software patent" lawsuits is making its way
across the American software industry.  With the 40,000+ software patents
to be awarded in this decade providing countless opportunites for such
actions, expect what follows to be the equivalent of the annual "hay fever"
distraction of being forced to pay money to treat just the symptoms.


    The first case is the so-called Freeny/EData patent that has been talked
about much in the press, mostly because its stock price seems to be rising
each time it makes some announcement.  In July of 1985, Charles Freeny was
issued a patent, 4,528,643, "System for reproducing information in material
objects at a point of sale location".  A simplified view of the patent is 
that it claims secure electronic commerce, which given the endless talk about
secure electronic commerce on the Internet today, makes it not surprising that
the current patent holders are sending out infringement notification packets
to 70,000 companies.


    I have expressed my skepticism for this patent at least a few times in
print, because like many of the 40,000+ software patents to be issued in the
1990's, the Freeny patent cites no non-patent prior art, even though before
the effective priority date of January 1982, there was much written about
electronic commerce and/or encrypting business communications (providing
grounds for an obviousness argument, if not a lack of novelty argument). Had
those materials been in the hands of the examiner, I am sure a different set
of claims would have issued.  Interestingly, one of the early licensees (which
EData touts all the time) was IBM, even though I think IBM could do some real
damage to this patent with IBM's own prior art, since in the 1970's, IBM did
a fair amount of work for industry and the government with regards to securing
network business communications.


    The patent has been reassigned five times, and expired/reinstated once,
and is currently owned by EData (formerly Interactive Gift Express).  A few
days ago, I received an infringement notification packet in the mail from
EData, very pretty (almost as pretty as Lemelson's CDROM he sends out to
notify people infringing his patents).  EData is offering "amnesty" (they
even include a dictionary definition of the word in their materials) to
potential infringers before they start suing.  They are going after three
Internet groups: content providers, resellers and service providers (what's
left?).  In the words of Arnold Friedlich, president of EData,


        "Part of our marketing strategy was to sue everybody and get
        noticed," Freilich said.  "Well, we went ahead and sued, and
        everyone now knows that we're very serious about defending our
        claims."


In mid-April, the price of EData stock was about $1.50, though by mid May
after a flurry of announcements, the stock was over $11.00.  Good notice.


    The interesting part is a licensing agreement they include which has a
table of payment schedules, a table that illustrates the "taxes" everyone
is going to have to pay as questionable software patents continue to issue
and provide a "license to sue".  Here is the table, with dollar figures
slightly rounded to fit on an ASCII screen:




                       Initial and renewal limited license fees ($)
  Company
Net sales($)      1996     1997     1998     1999     2000     2001     2002




<   25,000          500      550      600      650      700      750      800
<   50,000         1000     1100     1200     1300     1400     1500     1600
<  100,000         2500     2750     3000     3250     3500     3750     4000
<  200,000         4000     4400     4800     5200     5600     6000     6400
<  400,000         7500     8250     9000     9750    10500    11250    12000
<  750,000        10000    11000    12000    13000    14000    15000    16000
<1,500,000        15000    16500    18000    19500    21000    22500    24000
<3,000,000        25000    27500    30000    32500    35000    37500    40000
<5,000,000        40000    44000    48000    52000    56000    60000    64000
> 5,000,000                     subject to individual negotiation




Interesting fee levels - enough to make some good money - but not enough to
make it worth while for any one company to fight the patent on its own. A few
small Internet companies I know are already paying a few thousands dollars a
year to avoid dealing with EData's license to sue - one less ad they could
place to promote their businesses.


These fees are lots of money each year, when multiplied by the number of
copycat actions as others with similar software patents from the late 1980's /
early 1990's see EData succeed and decide to pursue similar license fees.  It
would be a lot cheaper for the industry to honestly and sincerely fund the
development of a comprehensive prior art database and nip these potential
lawsuits before they can sprout (couldn't resist - I just had another shouting
match with a Software Patent Institute member who asked for help for free).


At some point the cost of doing so will be cheaper than dealing with all of
these lawsuits (and I have seen a good argument that the PTO's reexamination
of the Compton's patent was wrong, which if Compton's lawyers succeed with the
appeal, brings the Compton's patent back into the picture). Certainly it is
going to cost members of the Software Publishers Association more than the
$300,000 the SPA donated to the SPI without getting anything of value for SPA
members (who might want to ask the SPA why they gave their dues money away so
cavalierly).


The future?  A lot of people are going to be looking at how well EData
succeeds, because what EData is doing isn't that expensive (say a few million
dollars for the direct mail campaign, etc.) for a potential substantial
reward (how about getting $1,000 a year from 50,000 companies).  It is the
type of risk/reward ratio that is common on Wall Street, and could lead to a
investment pools being formed to invest in software patent lawsuits (as well
as growing number of insurance agencies offering software patent infringement
insurance at a more reasonable price, since current premiums are about $50,000
a year).  It is as risky as investing in commodities, and tens of billions are
being invested there every year.


I estimate that every month for the last three years at least one software
patent has issued that I would class in the same category as the Freeny
patent - potentially infringed by everyone on the Internet / overly broad
independent claims / little-to-no non-patent prior art submitted.  As people
start making money on the Internet, expect these patents to start being
asserted.  Consider that for interactive television, another potentially
lucrative area, there are over 60 issued patents and 180 pending patents,
most for putting menus on your television screen.


So multiply the potential number of such patents, by the potential licensee
annual fees that will be sought plus annual insurance fees, and you have a
pretty good idea of what funds you might want to consider to set aside in an
escrow account to be prepared for what is inevitable.  Maybe the SPA should
write a spreadsheet to help its members calculate the escrow amount.


By the way, anyone using 3D pie charts on their computers, you are probably
next to be sued (but don't pay more than $3000, my fee to help you kill the
recently issued patent :-).  Either that, or Oracle's recent but questionable
parallel processing of electronic mail patent.  What's the Blondie song?
Right, "One way or another, we're going to get you ......" :-)


Greg Aharonian