[LWN Logo]
Date: Tue, 19 Dec 2000 11:37:26 -0500 (EST)
From: Gregory Aharonian <srctran@world.std.com>
To: patent-news@world.std.com
Subject: PATNEWS: Prior art for the British Telecom hyperlink patent

!20001219  Prior art for the British Telecom hyperlink patent

But first, newswires yesterday report that a federal jury has ordered
Boston Scientific to pay Johnson & Johnson $324 million for infringing
JJ's stent patents - $253 million for lost sales and $71 million as a
20% royalty on past sales.  Boston Scientific will a) ask a judge to
set aside the verdict, b) present evidence that the patents were 
obtained "improperly", and c) appeal to the CAFC.


Yesterday, I sent out a PATNEWS that mentioned how British Telecom is
shamelessly suing Prodigy over BT's hyperlink patent.  That's all I was
going to report on the matter, until I got a response from a lawyer from
one of the big London law firms, apparently a firm that does a lot of
work for BT, that I think is an insult:   "Greg, fuck off you git!".
After turning down this generous offer and kicking him off the list,
while eating Hamm and cleaning my plates with Suds, I decided to have
some more fun at BT's expense.  What follows is some prior art, a letter
from an editor of a Linux Journal castigating both British Telecom and
BT's lawyers (Kenyon and Kenyon), followed by my earlier analysis of BT's

First the prior art.  To critics of the BT patent, keep in mind one thing.
The priority date of their US patent is July 1976.  It doesn't matter that
the patent issued in 1989.  Any prior art has to be before July 1976, and
to be honest, there isn't much, and what there is, part of it is BT's own
work with hyperlinks for their teletext and videotext systems.  You have
to keep in mind that BT was one of the first companies to get involved
with and commercialize hyperlinks.

But the key phrase is "one of the first", not "the first".  What will come
to haunt BT's efforts are four papers published in the 1960s:

        A hypertext editing system for the IBM 360 (Nelson)
        1968 Univ. Illinois COnference on Computer Graphics, 291
            "Any text structures may be interconnected (LINKED)
            in arbitrary ways, and the user may jump along
            connections in this linkage structure."

        A research center for augmenting human intellect (Englebart)
        1968 AFIPS Fall Joint Computer Conference, 395
            "User files are organized as hierarhical structures of
            data entities, each composed of arbitrary combinations
            of text and figures. ..... with provisions of arbitrary
            CROSS-REFERENCING among the elements of a hierarchy."
            This system made use of a mouse.  To see a video:

        AUTONOTE: a personal information storage and retrieval system
        1969 ACM National Conference, 67
            "... He has available mechanism for deleting, replacing,
             LINKING and hierarchically organizing text items."

        SHOEBOX - a personal file handling system for technical data
        1970 AFIPS Fall Joint Computer Conference, 535
            cites Engelbart's and Nelson's papers - had a nice
            search facility.

(Speaking of shoeboxes, if someone at the PTO wants a copy of these papers
for the examiners shoeboxes, let me know.)

LINKED.  CROSS-REFERENCED.  LINKING.  All in the 1960s.  If BT is not
aware of these papers, it should fire whoever was involved with assessing
this patent.  And if BT is aware of these patents, they should be spanked
for then suing anyone.  At best, BT's patent in light of these papers
will be narrowed to protect what the patent was meant to protect - BT's
specific implementation of hypertext in their 1980s teletext and videotext

Once again, I will argue something that makes me an outcast: that there
is nothing wrong about patenting software (which upsets the free software
crowds) and much wrong about the examining of software patents (which
upsets everyone else - seriously there has been zero improvement in the
quality of issued software patents because no one really gives a damn).
The same will hold true for business method patents.  The only reason that
so many bad business method patents will issue is because Rule 56 is
pathetic, and no one who is whining and crying like a baby has the guts
to publicly support patent fee increases or toughening Rule 56.

So I ask you, who is the git?

Greg Aharonian
Internet Patent News Service



          -= Open Letter to Mr. Charles J. Roesslein, CEO, Prodigy =-

                          by Don Marti <dmarti@ssc.com>

To:  Charles J. Roesslein, CEO, Prodigy 

Dear Mr. Roesslein, 

As a company, you just got infested with worms. Parasites trying to nourish
themselves from your success -- software patenteers.  If any of us other
Internet builders and users can offer you some assistance fighting them,
please let us know. 

Since I work for a company that makes hyperlinks among other things, and
since I make (burnallgifs.org) some links in my spare time, I naturally
got worried about the fact that British Telecom is actually suing you over
their ridiculous claim that they, not Vannevar Bush, Ted Nelson, Doug
Engelbart, or anybody else invented hyperlinks. Could this be the end
of the Web as we know it? 

But then I started to relax a little. You see, BT's law firm is Kenyon
and Kenyon, a protection racket that makes its living shaking companies
down for royalties on bogus patents. Bottom feeders. They would probably
be sending out green card spam if they knew how to work a computer. 

The Internet community has beaten Kenyon & Kenyon harassment before, and
we can do it again. Do a web search on "+CueCat +Linux" and you'll find
that they're the firm that tried to intimidate a bunch of Linux hackers
into ceasing, desisting, and otherwise caving in to legal threats over
independent software developed to make the CueCat work with Linux. As we
should all know by now, the Linux software that these latter-day Stamp
Taxers tried to stomp out now flies free over the land of the free and
the home of the brave. 

Brave. That's the key word here. If a few low-budget C programmers can
stand up and be counted, and assert their right to advance the common
computing goals of humanity, then you can too. Settlements don't buy you
peace in any time -- they just attract more parasites. Ever hear of a dog
with "worm", singular?  No matter what they offer, no matter how much
they threaten, fight. Your shareholders will thank you next year, when
the next batch of parasites leaves you alone. 

So, it's not fair for us professional and amateur hyperlink fans to urge
you to resist a harassing lawsuit and then leave you to do it by yourself.
So what do you need?  A well-publicized search for prior art to strike
their patent down?  Mozilla got help from the Internet community to beat
a similar bogus patent lawsuit (Wang claimed they invented bookmarks),
and maybe people can find some more good examples of systems that look
like today's hyperlinks a hell of a lot more than whatever old crap BT
is waving around looks like today's hyperlinks.

Or does Prodigy need any other help, for that matter -- a hand truck or
two for your file cartons full of prior art? Or a cool beverage?  Just
ask. The Internet stands with you on this one as long as you don't back
down. BT and Kenyon and Kenyon may think that they can make you pay up
first before they move on to bigger prey such as AT&T and AOL.  But
they're pulling on the essential wires of the Internet, and the rest of
us would be fools not to help defend it. What do you need? Don't be shy. 

When clueless companies start throwing "intellectual property" claims
around to suppress competitors or to extract money from innovators who
have left them in the dust, that's a threat to our economy and, when
they attack free communication protocols, it's a threat to our freedom.
In the long run, I hope that this case will make you as dedicated a
software patent reform advocate as I am. But for now, don't give the
bastards an inch and you'll get all the help you need. 

Don Marti 
Technical Editor, Linux Journal 
Copyright 2000 Specialized Systems Consultants, Inc. 

(GREG NOTE: I doubt that Kenyon & Kenyon cares that the open source world
doesn't hold it in high regard.  Still they might want to consider some
spin control.)



20000621  British Telecom patent: too lachey, uninfringeable and invalid

    (Quick note: this week's Barron, 19 June 2000 issue, page 52, has a
short article about how Microsoft's 1200+ patents will be split if the
company is split.)

    Now that everyone inside and outside of British Telecom has had their
fantasies of cashing in on BT's patent that is being asserted against
Internet Service Providers, and the folks at Aurigin have nominated the
BT patent for the "'Rembrandt' in the Attic" of the year award, and the
press have made long term plans for something new to write about, let's
hear from Greg's peanut gallery on why this patent's strength is up there
with wet spaghetti.  In short, this patent probably better qualifies for
the "'Dogs Playing Poker' in the Basement" award.

    How will this patent (4873662, Information handling system and terminal
apparatus therefor) be attacked?  Three modes: laches (possibly but
unlikely), noninfringement (probably) and invalidity (more probably).
I don't know what British Telecom was doing in its past few years
analysis of this patent, but it has major problems.

    First laches, which is lawyer lingo for the situation sometimes that
if you wait too long to assert your patent, you can't.  BT's patent issued
in 1989, and the Internet took off commercially in the 1995 time period,
so filing lawsuits in the year 2000 does indicate an inordinate amount of
time to assert the patent.  However, this is an unlikely attack mode
against the patent, in view of the comment of one lawyer who gets PATNEWS:


      Yes, digging really old patents out of the closet to belatedly
      sue someone has become a lucrative "lawyer cottage industry",
      because of the above [the right to sue six years after a patent
      has expired for past damages] and also because: (1) the U.S. has
      never [stupidly, in my view] adopted a normal, real, "statute of
      limitations" or "statute of repose" law to prevent it [as there
      is for almost any other "stale claims" tort suit], and (2) the CAFC
      has made patent suit "laches" into a mere weak and case-confused
      "double bursting bubble", with a presumption of laches normally
      arising only after 6 years of KNOWING about the infringement before
      bringing suit.  "Laches" only prevents collection of BACK-damages

    So if laches is out, let's turn to non-infringement.  Is there anyone
out there actually infringing the BT patent's claim?  Using a simple
analysis, the answer is NO.  First, the view of another lawyer reader


    I read about this today as well and my first thought was that, if the
    patent is actually litigated, it is almost certain to make law in the
    means-plus-function claim interpretation arena as all of the elements
    are recited in terms of Section 112 means-plus-function format and
    there is literally no description of any equivalent embodiments other
    than the two preferred embodiments of dumb ascii terminals option A
    (having no local memory for storing the hidden addresses - claim 1 -
    which cannot be infringed by anything other than a dumb terminal) and
    option B (having a local memory for storing the hidden addresses in
    response to digital keyed input - claim 3).

    The question for the CAFC will be whether a PC today having more
    computational horsepower than the central computer described in
    this patent can be the equivalent of the dumb ascii terminal that
    is described as the "remote terminal means".  My two cents would
    be that BT will have a tough time convincing a court that a PC of
    today is the equivalent structure to a dumb ascii terminal with
    two memories, no processor and only a memory control unit to manage
    the two memories.

Sounds like bad news for BT.  Additionally, the independent claims are
written in the dangerous client/server mode, for which no one entity
infringes.  For example, claim 1 reads as:

    A digital information storage, retrieval and display system comprising:

    - a central computer means .................

    - plural remote terminal means .................

    - ..... [central computer means interacting with terminals] ......

Independent claim 3 has the same structure, and claims 2 and 4 are trivial
dependent claims off of claims 1 and 3.  But as has been questioned about
(Internet) client/server patent claims, who actually infringes claim 1/3?
Internet Service Providers as entities don't infringe the remote terminal
means clause of the claim (their customers, being the remote terminals,
are separate legal entities).  Subscribers to ISPs while infringing the
remote terminal means clause, don't infringe the central computer means.

The suggestion for these type of split systems is to claim the central
computer/server and remote-terminal/client separately, so that infringement
can be asserted.  Ironically, the BT patent does this, but for the wrong
component.  The remaining claims, 5, 6 and 7, claim the terminal apparatus
and its interaction with the central computer, which certainly would be
difficult to assert against ISPs, and pointless to assert against people
with Internet accounts at the ISPs.

So after having to narrowly interpret these means-plus-function claims,
is there anyone infringing in its entirety whatever is left to be claimed?
And if there is anything left to be claimed after the claims analysis,
is what is left novel and unobvious in light of prior art not disclosed
to the PTO?  In short, is what is left a valid patent?  In particular,
is the concept of blocks of texts with pointers to other blocks of texts
and with formatting information for the texts, are these techniques
novel and unobvious?

Most likely, NO.  As many PATNEWS readers have pointed out, the British
Telecom patent looks a lot like the writings of Ted Nelson's Xanadu
hypertext project from the early 1970s (its history is documented at
www.xanadu.net), which easily qualifies as prior art.  What will haunt
British Telecom is a paper not cited in the issued patent written by
Nelson and others in 1969, but indirectly referred to by others to the
extent that it should have been found.  For example in Roy Rada's 1991
book "Hypertext: from text to expertext", he writes:

      In the late 1960s, the Hypertext Editing System was developed
      on a mainframe computer.  At the time, the normal technology
      for editing on mainframes was batch cards.  The Hypertext
      Editing System supported branching text and automatically
      arranged branches into menus.  Authors could specify which
      branches to follow when printing was to occur.

      The Hypertext Editing System failed in the marketplace.  In
      1968 the Hypertext Editing System was demonstrated to staff
      at two major publishing corporations, who staff felt, however,
      that the Hypertext Editing System was too complex.  The idea
      of sitting being a computer terminal and authoring and 
      editing was more than the managers at that time were willing
      to believe.

While the Hypertext Editing System failed in the marketplace, it was around
long enough for at least one paper to be published that survives until
today.  In early April of 1969, the University of Illinois (ironically
the home in the 1990s of the first Web browser, Mosaic) held the "Second
University of Illinois Conference on Computer Graphics".  One of the papers
was presented by people from the hypertext group at Brown University (the
future home of another hypertext system, Dynabook), one of the persons
presenting the paper being Ted Nelson.

The paper is titled "A hypertext editing system for the 360", and
described a system with a IBM 360/50 mainframe running a hypertext
system accessible by IBM 2250 dumb terminals.  I first paraphrase the
paper's description of links:

      Areas of text may be connected in two ways: by links and/or
      by branches.  A link goes from a point of departure to an
      entrance point in another, or the same, text area.  Links
      are optional paths embedded in the text.  A point within an
      area may also be given a name (label) and later summoned by
      name to the screen.  The user goes from area to area by
      "travelling" ("jumping") via a link or a branch, or by
      "getting" a label.  The text is repositioned on the screen,
      so that an entrance-point is always at the upper left corner
      of the screen, and with as many text lines following it as
      will fit on the screen.

Seems to me that this completely anticipates BT's claim language such as
".... in which plural bodies of information are stored at respectively
corresponding locations, each of which locations is designated by a
predetermined address therein by means of which a block can be selected
....".  In this 1969 paper, we also read:

      He may "format" the text for ease of reading on the screen
      or for final output, specifying margins, number of columns,
      paragraphs, indentations, types of headings, underscoring, etc.
      ...... The formatting program permits the user to assign format
      codes to characters and strings within the text.

which seems to anticipate words from the BT patent's abstract: " .....
The second part of the block could alternatively influence the format
and/or color of the display at the terminal."   Admittedly the block
structure of the BT patent, and the text area structure of the 1969
paper are different, but not different enough to escape being obvious.
And in terms of applications, the 1969 paper anticipated the future plague
of the Internet:

      Explainers may be used to tell what kind of area a link or
      branch leads to, and so act as "teasers" or ADVERTISEMENTS.


So under this combined assault of laches, noninfringement and invalidity,
my guess is that British Telecom is going to have a hard time asserting
this patent against rich deep-pocketed ISPs who can afford to pay the
best law firms to take these types of arguments and available prior art,
and make much stronger versions of the above analysis.