|                                                     |
         |       PROGRAMMING FREEDOM  -  online edition        |
         |                                                     |
         |             March 1993  -==-  Number 7              |
         |                                                     |
         |           The Electronic Newsletter of              |
         |        The League for Programming Freedom           |
         | 1 Kendall Sq #143, POBox #9171, Cambridge MA 02139  |
	 |          Send email to: lpf@uunet.uu.net            |
         |        Voicemail phone number: 617-621-7084         |  
         |     [Notice that the phone number has changed!]     |
         |   Leave your message and we'll return your call.    |
         |  Editor: Spike R. MacPhee (spiker@prep.ai.mit.edu)  |
         |           Managing Editor: Gordon Schantz           |
         |           Layout/Design: Susan Hofstader            |
         |     Reproduction of Programming Freedom via all     |
         |            electronic media is encouraged.          |
         |     To reproduce an individual signed article       |
         |       please contact the author for permission.     |


The LPF at COMDEX FALL 92........................Chris Hofstader
Case Study: 
Losing Market Share Because of Patents...........Bob Sutterfield
Excerpt from "A Statement of Principle"...........Bruce Sterling
Another Allegedly Trivial Patent Case: 
Stac Vs. Microsoft ................................Ross Williams
League Email adresses...........................................
Purify - An Example of Patent Entrapment........................
Legal News: Sega Vs. Accolade, and Atari Vs.Nintendo Court Cases
IBM vs. Australian Commissioner of Patents......................
LPF Boutique....................................................
Chris Hofstader, LPF Board member

     LPF members in attendance: Jack Larsen, Gordon Schantz,
Randy Stankey, Mike Robert, Dean Anderson, Adam Richter, and
     This was the first time that the LPF hosted a booth at a
trade show focusing on the end user rather than programmers or
academics. It is also the largest convention of any kind on the
     Over 5000 of the people who visited our booth in the Riviera
took a position paper!  This means that the people in our booth
spoke to over 1000 people per day. 
     But more remarkable than the number of people to whom we
were able to speak was the overwhelmingly positive reaction that
we received.  Talking amongst ourselves we could not remember
more than 20 people who reacted unfavorably to our position.  We
could all remember hundreds who reacted extremely positively and
suggested that they would likely join the LPF.  Remarkably, the
most supportive reactions came from leaders (Presidents and such)
of small companies in the US and abroad. The LPF is very excited
about future cooperation with this segment of the industry.
     Many members of the press stopped by including: a reporter
from Reuters, AP and many other non industry and industry
magazines and news services.
     Notables at COMDEX Fall '92 included the following:

*    Phillipe Kahn, President of Borland, was seen wearing a
"fanged apple" badge and expressed support for our efforts.

*    At the Party/Rock concert sponsored by MicroGrafix everybody
working in the Microsoft and Borland booths was wearing a fanged

*    Author John Dvorak committed to doing "something big" (to be
determined later) with us in 1993.

     This year, though, the big story is small.  Small companies
seem to love our position.
Case Study: Losing Market Share Because of Patents 
Bob Sutterfield (bob@roughy.MorningStar.Com)

     The specific subject: PPP FCS suboption negotiation
     The Point to Point Protocol (specified in RFCs 1331, 1332,
and others) is a product of the Internet Engineering Task Force's
Point to Point Working Group.  It is the successor to SLIP (RFC
1055), and provides many useful facilities that were lacking in
SLIP.  One of the purposes of PPP was to create a single standard
IETF-blessed inter-router framing protocol, so that routers from
different manufacturers would easily be able to interoperate.
     One facility that's a standard (non-optional) part of PPP is
that each frame carry a FCS (CRC Frame Check Sequence) field.  By
default, the FCS is 16 bits long, though peers may agree to use
32-bit FCS for improved error detection.  The existing method for
that agreement is for the administrators setting up the link to
simply modify their configurations appropriately, by hand.
     This "prior  agreement" method of verbal negotiation is
inconsistent with PPP's goal of providing in-protocol methods to
negotiate every variable that affects the connection (frame size,
network-level addresses, compression methods, authentication,
etc.)  But there's a problem in that frames with one size FCS
will be rejected by a PPP that's expecting another size FCS, so
the peers will never be able to negotiate their way to agreement.
     So in December of 1990, Arthur Harvey of DEC published a
Draft RFC describing a clever negotiation scheme: It's possible
to create a valid 48-bit FCS for a frame, such that the first 16
bits will be a valid 16-bit FCS, and the first 32 bits will be a
valid 32-bit FCS. This will allow the negotiation packets
themselves to be exchanged without being discarded because they
have a bad FCS value, since the sender's initial FCS size didn't
match the receiver's.
     In late 1991, on the IETF PPP WG mailing list, a co-author
of the Draft RFC mentioned that DEC has applied for a patent on
this technique.  Of course, he said that they plan to offer
licensing terms, but the fact remains that the technique will
likely soon be patented.  This made PPP developers (like our
company) nervous, because we don't know what those terms might
be.  Now in late 1992, DEC has published their terms: $5000 for
the right to use their negotiation technique.  We've implemented
both 16-bit and 32-bit FCS in our product, but we may never
implement DEC's negotiation scheme.
     Fortunately, during 1992, several PPP developers proposed
alternatives to DEC's negotiation scheme.  Any of these will make
the optional improved error detection more conveniently available
to users of wide-area networking hardware and software.  Since
DEC wants to charge a license fee, it's unlikely that any other
vendors, or authors of free software, will implement it.
     Some more points of possible interest: The Draft RFC that
was published in December 1990 contains no mention that a patent
is pending on the negotiation technique.  I have suggested that
the author include such a warning for unsuspecting potential
implementors like ourselves.  If we had gotten around to
including DEC's negotiation scheme in our product during the
period when DEC hadn't yet told the PPP community that it was
attempting to patent the technique, we would have been
retroactively and unexpectedly dunned for license fees. 
Standards bodies should be particularly careful that their
contributors don't attempt to use this sort of strategy.
     Also, in response to the 1991 DEC announcement of the
pending patent, a good-hearted PPP developer made his 16/32-bit
FCS negotiation code available for FTP, and placed the
implementation in the public domain. So even the 16/32-bit
negotiation implementation is free (as are implementations of the
rest of PPP); only the spectre of the patent is stopping
developers from providing the DEC-style negotiation facility to
their users.  DEC's negotiation scheme may turn out to be a
DEC-specific enhancement to PPP, available only in DEC products.
Everyone else in the PPP community will use some other
negotiation method.  In this case, the good news is that DEC's
patent application has backfired: it gave DEC a black eye, and
ensured that DEC's scheme would not be commonly implemented.
From "A Statement of Principle" by Bruce Sterling  (Science
Fiction Eye #10):
     "There's something wrong with the Information Society. 
There's something wrong with the idea that 'information' is a
commodity like a desk or a chair.  There's something wrong with
patenting software algorithms.  There's something direly mean
spirited and ungenerous about inventing a language and then
renting it out to other people to speak.  There's something
unprecedented and sinister in this process of creeping
commodification of data and knowledge.  A computer is something
too close to the human brain for me to rest entirely content with
someone patenting or copyrighting the process of its thought.
There's something sick and unworkable about an economic system
which has already spewed forth such a vast black market.  I don't
think democracy will thrive in a milieu where vast empires of
data are encrypted, restricted, proprietary, confidential, top
secret, and sensitive.  I fear for the stability of a society
that builds sand castles out of databits and tries to stop a
real-world tide with royal commands."
>Bruce Sterling is a so-called "cyberpunk" (among other
descriptions) science-fiction writer, and an LPF member.
Another Allegedly Trivial Patent Case:
Stac v.. Microsoft
Ross Williams, ross@spam.adelaide.edu.au
     As many of you may know by now, Microsoft is being sued by
Stac  Electronics, who claim infringement of a data compression
software  patent. The patent is US Patent #4,701,745, the
"Waterworth" patent,  and has a priority date of 6 March 1985.
(Its abstract and claims sections are available upon request from
the LPF.)
     MS-DOS 6, due to be released in the first half of 1993, is
the target  of the Stac suit. MS-DOS 6 contains transparent disk
compression, a  key component of which is the data compression
algorithm under dispute. 
     It's not hard to work out why this patent lawsuit exists. By 
incorporating transparent disk compression into MS-DOS, Microsoft
will, for the first time, be competing directly with third-party 
transparent disk compression products, and this is likely to
precipitate a shake-out in the niche. This filtering down from
third party products to the operating system is a natural
progression in technological development, and should be applauded
(encryption is probably next in line). However, it must be
extremely threatening to companies such as Stac Electronics who
live in this transitory niche. 
     The patent turns out to be rather significant in a number of
ways. Not only is it a very early software patent (1985), but it
actually seems to cover most of the content of a LATER patent --
the "Gibson and  Graybill" patent #5,049,881 (registered in 1990)
-- so recently fought over by Intersecting Concepts and AdStor.
This makes the Waterworth patent the king pin of a small tree of
     Having myself re-invented and published in 1989-1991 an
algorithm (my LZRW1 algorithm) very similar to that described in
the Waterworth patent, I am completely convinced of the patent's
triviality. In a nutshell, the patent covers the use of hashing
to find matches in a sliding input window. This is a fundamental
and obvious technique in data compression and it is ludicrous
that it should be anywhere but in the public domain. In fact it
is ludicrous that it should have ever even existed as a piece of
intellectual property. 
     The fact that a company as large as Microsoft is defendant
in this case means that the stakes are high. If Microsoft wins
then the whole tree of patents will fall. If Microsoft loses,
then it is unlikely that the patent will be challenged again and,
as a consequence, it will be massively strengthened. I believe
that it is in the interests of all those involved with data
compression (except of course you know who) to see that this
patent is toppled. 
     Because of my close involvement with these kinds of
algorithms, and my sympathy with their cause in this instance,
Microsoft is employing me on a contract basis to assist with the
case. Currently I am collecting prior art. If you know of any
prior art, or have any idea of where it might be obtained, I
would be most grateful if you could contact me at my email
LPF email Lists.
     These lists are for LPF members only, although you may, of
course, redistribute postings to your friends in the hopes of
getting them to understand the LPF positions and possibly
actively support the LPF by joining. 
     The moderated mailing list: 
league-activists@prep.ai.mit.edu and its two sub-lists:
league-activists-boston@prep.ai.mit.edu and
league-activists-remote@prep.ai.mit.edu should be used only for
members' requests for assistance in league projects, local or
nationally, or for announcements from LPF.
     These lists are filtered by a moderator to:
*    insure this use;
*    minimize the number of messages;
*    remove items meant for the list's -request address;
*    forward items that should have been sent to another list.
     There may be a delay of up to 3 days for your message to be
sent on L-act, so plan ahead for volunteer requests.
     League-tactics@prep.ai.mit.edu is for discussion of LPF
directions and is not moderated. If you want to subscribe, change
your eddress (email address), or be removed from either list,
please use: league-activists-request@prep.ai.mit.edu or
     General questions about the LPF, and administrative
questions about your membership or your email copy of the
newsletter should still go to: lpf@uunet.uu.net.

An example of patent entrapment 
in a techniques-patented program paper at a Usenix conference
     A recent Usenix conference had a paper which describes
techniques that are being patented by the authors.  This was
revealed by the fact that listings of output from the program
showed the words "patents pending".  The program is called
"purify" and it checks for storage leaks.
     There is nothing brilliant about the program--it is a
different arrangement of commonplace ideas.
     In effect, the paper is just a sales presentation, since
anyone who tried to use the ideas learned from the paper would be
     The LPF has suggested to the Usenix board that they prevent
such occurrences in the future by insisting that all authors sign
a statement that they do not know of any attempt to patent what
they are describing and won't patent it themselves.
IBM v. Australian Commissioner of Patents (PC Week (Australian),
February 3, 1993, p 22):
     A software patent crisis? It's patently obvious!
     A little-noticed court case, now concluded, has paved the
way for computer programs to be patented in Australia. This will
have enormous implications for the our [sic] IT industry because
it will mean that software developers will have to be wary when
incorporating features into programs. It will also strengthen the
hands of big computer companies like IBM, which now employs
scores of patent attorneys to obtain money from small computer
companies, if necessary by suing them.
     Back in 1991, IBM Australia applied for a patent on a method
of making smooth curves on computer screens. The commissioner of
patents rejected the application, arguing that the claimed
invention was nothing more than a mathematical formula.  But
Federal Court judge Burchett held on December 13 1991 that the
application was not for an abstract formula but for a new and
inventive application - and was therefore patentable.
     The commissioner of patents decided to appeal the decision
to the full Federal Court, but in August 1992 the appeal was
withdrawn, so that the original judgement has been upheld.
     IBM versus Commissioner of Patents brings Australian law
much closer to that of the US where the US Patent Office has been
issuing more and more patents for programs - and even small bits
of programs - much to the consternation of some elements of the
software industry.
>Tom Forester, senior lecturer, School of Computing & Information
Technology at Griffith University, Brisbane.

[Excerpted from Technology Law Bulletin, January 1993, article by
Lucash, Gesmer & Updegrove,  sent by Jonathan  ]:
Reverse Engineering Software To Achieve 
Compatibility Is Fair Use
     Sega Enterprises Ltd. v. Accolade, Inc.: Accolade
disassembled Sega's console ROM to discover the sequence of 25
bytes required as a "security code" by the Sega Genesis video
game.  Sega claimed that Accolade had violated their copyright on
the console ROM software, but the 9th US Circuit Court of Appeals
held that disassembly of a copyrighted work is fair use if
"disassembly provides the only means
of access to those elements of the code not protected by
copyright and the copier has a legitimate reason for seeking such
access."  Sega was willing to license the code to Accolade only
under an 'exclusive business relationship' [presumably, only to
market their games on Sega systems], which the court found was
unreasonable.  Furthermore, the 25 bytes were found to be
unprotected functional elements of the program.
     Atari Games Corp. v. Nintendo of America, Inc.: Atari also
reverse-engineered the Nintendo console security system, but part
of the means they used to do this was to obtain a copy of the
(copyrighted) program which generated the security codes from the
Copyright Office under false pretenses.  Atari was thus held to
have infringed Nintendo's copyright based on the fact that there
were many different programs which would generate a password for
Nintendo consoles.
LPF Boutique: 
Materials Available from the League
Please send orders to: 
League for Programming Freedom
One Kendall Square #143/PO Box 9171
Cambridge MA 02139. 
     We don't take credit cards yet, but do take US currency in
checks, money orders, or any of the brands of Travellers Checks.
(Please don't send cash in the mail).
Buttons: We have reprinted the famous "fanged apple" buttons. 
These buttons show the symbol of Apple computer with an alien
snake's body and face.  You can buy buttons by mail from the
League, for $2 each, in quantities of at least three.  We give
out buttons at events, but ask for a donation.
Stickers: We also have stickers showing Liberty Empowering the
Programmer, with the League's name and address. You can order
stickers by mail from the League at $5 for 10 stickers; for
larger orders, phone us to discuss a price.  We hand them out
free when it is convenient, such as at our events, but since
mailing packages to individuals costs money, we want to make it
an opportunity to raise funds.
     Post stickers at eye level and separated from other posted
articles, to make them easy to see.  The stickers are not made to
survive rain.
Liberty Postcards: We also have postcards showing Liberty
Empowering the Programmer, with the League's name and address. 
$5.00 for 10; call for volume discount..
Large Liberty Posters: We have a few posters with the same image
that is on the stickers, approximately 2.5 ft by 1.5 ft.  They
are $4 each and $4 total shipping and handling in the US for the
first one to five posters, and $2 shipping/handling for each
additional five.  We also have a Postscript file of the Liberty
image available at the FTP sites under the name:
/pub/lpf/articles/liberty.ps.Z for you to download; it's a 100k
Coffee Mugs: Our coffee mugs have the Fanged Apple design in full
color on one side and "League for Programming Freedom" on the
other.  They hold twelve ounces and are microwave safe.  You can
order a mug for $15, nonmembers $17, plus $3.00 shipping and
handling.  They are now in stock.  Note the price increase.
T-Shirts: Michael Ernst has produced t-shirts with Liberty and
"League for Programming Freedom" on the front and "Innovate,
Don't Litigate" on the back.  (The back slogan will change from
time to time.)  You can order shirts by mail from the League for
$10, nonmembers $12, plus $2 for shipping and handling. 
Available colors are yellow, light blue and ecru; if you specify
a color, we will assume you would rather have another color than
no shirt.  If you want a chosen color or nothing, say so
explicitly.  Please specify the shirt size!  (M, L or XL.)  We
are sold out of XL shirts with this back-slogan.
     We have printed the next version of the LPF t-shirt.  The
new back-slogan is "You'll pay for this" with an XORed cursor
over the word "this", and "League for Programming Freedom"
underneath.  The front is the same as the older shirt, and the
colors are yellow, light blue, and off-white in M, L, and XL
Position Papers and Memberships:  We will send anyone a copy of
the League position papers.  If you want other copies to hand out
at an event, we'll send you as many as you need.  Please discuss
your plans with us.  One-year memberships are $42 for
professionals, $10.50 for students, and $21 for others. The dues
are $100 for an institution with up to three employees, $250 for
an institution with four to nine employees, and $500 for an
institution with ten or more employees. For $5000, an institution
can be a sponsor rather than a member.
League Papers On-line:  You can retrieve LPF written materials in
TeXinfo format by anonymous ftp from prep.ai.mit.edu in the
directory /pub/lpf.  These include the position papers, all back
issues of our newsletter Programming Freedom membership form,
handouts, friends of the court briefs, and articles about the
LPF's issues of concern. In addition to the above, Joe Wells has
PostScript, DVI, plain text, and Info format versions of the
papers "Against User Interface Copyright" (look-and-feel) and
"Against Software Patents" (patents) available for FTP from the
location: cs.bu.edu:pub/jbw/lpf/
League Video Cassettes:  We have a four-hour video tape of two of
Richard Stallman's speeches for the LPF.  If you'd like to give
LPF speeches, we can send you a copy of this tape to give you an
example to learn from.  If you'd like copies for another purpose,
we can send them for $20 each (includes $4 shipping and
handling.)  They are now available in VHS/NTSC format only.
The LPF  Needs Volunteers

The LPF has been becoming increasingly active as we grow and
mature as an organization.  It seems however that the more the
LPF accomplishes, the more we need to do.  As most of you know,
the LPF only has one part time administrative person and one
other paid staffer to work on our publications.  The rest of our
work, from our directors on down to people who hand out stickers
at trade shows, is all done by volunteers.  

The LPF 's bi-monthly publication, which you are holding in your
hand, takes a lot of work to produce.  This can be done by our
staffers but the content must come from you. We invite you to
submit anything you think may be of interest to the
membership--opinion, case studies, current events,
absurd-patent-examples and other relevant humor, etc.  If we
receive it by the deadline and our editors deem it worthy, you
may see in one one of our upcoming issues:

April 28      draft articles due
June  1       newsletter mailed

June  28      draft articles due
August 1      newsletter mailed

August  28    draft articles due
October 1     newsletter mailed

October 28    draft articles due
December 1    newsletter mailed

December 28   draft articles due
February 1    newsletter mailed