The closely watched case of the Freeny electronic-commerce patent is gathering steam (see TBTF for 1996-04-07 [1] and the recently updated timeline [2]). Having sued 43 companies and sent letters of infringement to a further 139, E-DATA, the patent holder, began sending letters offering "amnesty" to 75,000 software authors and vendors who agree to settle on given terms. And on June 21 the dispute saw its first day in court.

At the same time E-DATA took out ads in the media [2a] (I saw one in IDG's Webmaster magazine). It carries a simple message. A picture of a carrot and a stick is labeled "Carrot" and "Stick," just in case you missed the point. The text reads: "Your choice. E-DATA Corporation announces Amnesty. U.S. Patent # 4,528,643 and its foreign counterparts, held by E-DATA Corporation, cover on-demand electronic distribution. If you infringe on its claims, you are violating E-DATA's intellectual property rights." The amnesty offer runs from June 1 to Aug. 31.

How far down into the software barrel does a mailing list of 75,000 get you? One who received a package was Dan Veaner of EmmaSoft Software [3], maker of the shareware program Electronic Postal Window, which does lookup of postal rates. The program has a facility for entering a code sent to you after you pay the shareware fee to stop the annoying "Please pay up" messages. Numerous shareware programs have a similar feature. E-DATA claims this feature infringes the Freeny patent. Did E-DATA send a letter to every shareware author who employs this technique? It's possible they did; it's likely they tried.

The proposed royalty schedule [4] in the amnesty letter represents a sliding scale from 5% of annual revenues for the smallest companies down to 1% for larger ones. There is no provision for a one-time, paid-up license. These numbers are quite high. In comparison, consider the royalty awarded to the modem maker Hayes Microcomputer Products for licensing the Hayes command set -- a facility required by essentially all modems in the market. A jury trial resulted in a royalty of 1.75%. E-DATA's attorney, David Fink, has been quoted characterizing the proposed fees as modest, saying that a fee in the neighborhood of 15% - 20% would be fair.

The Federal judge who held the June 21 conference, Barbara Jones, is new to the year-old case. Meeting with attorneys for E-DATA and for more than a dozen defendant companies, she required E-DATA to give each defendant a list of which products may be infringing the patent and which of the patent's 56 claims those products may be infringing. This is modest good news for the defendants, because E-DATA wants to keep their claims as broad and nonspecific as possible, the better to frighten possibly infringing companies into settling.

The judge indicated that she will probably hold a so-called "Markman hearing" as early as late August. These hearings arose from recent Federal court judgements, unanimously upheld by the Supreme Court, that judges and not juries are the appropriate vehicle for determining a patent's scope (that is, the meaning of its claims). The result of Markman hearings over the last year has been to shorten patent trials.

For an excellent summary of the case see Information Law Alert's coverage [5]. This publication also carried the best reportage I've found of the 6/21 hearing at [6], jocularly titled "25 lawyers and a judge." The Netly News published a readable article [7] about the patent on 6/6. It contains one substantial error, however -- the claim that Compuserve settled with E-DATA. The list [8] of settlements to date is: Adobe Systems, CD-Max, First Virtual, I.B.M., Index Stock Photography, KidSoft, MicroPatent, Monotype, and VocalTec.

[1] <http://www.tbtf.com/archive/1996-04-07.html>
[2a] <http://www.ramresearch.com/crdflash/e_data.html>
[2] <http://www.tbtf.com/resource/freeny-timeline.html>
[3] <http://members.aol.com/emmasoft/>
[4] <http://infolawalert.com/stories/053196a_licenses.html>
[5] <http://infolawalert.com/stories/053196a.html>
[6] <http://infolawalert.com/articles/960628_e-data_hearing.html>
[7] <http://pathfinder.com/Netly/daily/960606.html>
[8] <http://www.3wnet.com:80/corp/edata/family.html>

The U.S. Secretary of Commerce announced plans to appoint a panel to advise on implementing a "key-management infrastructure" [9]. This makes four times the Clinton administration has tried to push through a proposal requiring users to disclose their private keys to a government-certified escrow agent. The KMI "grandson of Clipper" proposal, first advanced by the White House in May [10], was widely criticized by members of Congress and the public.

The new proposal runs counter to the recent findings of a National Research Council study (summary at [11]; see also TBTF for 1996-05-31 [12]) that concluded it would be a mistake to continue "aggressive promotion" of key-escrow encryption. The Commerce Department's own Computer System Security and Privacy Advisory Board has endorsed the NRC report's conclusions.

[9] <http://www.epic.org/crypto/key_escrow/white_paper.html>
[10] <http://www.tbtf.com/archive/1996-05-20.html>
[11] <http://www2.nas.edu/cstbweb/28e2.html>
[12] <http://www.tbtf.com/archive/1996-05-31.html>

David Sol Bennahum <davidsol at panix dot com>, whom some of you may know from his MEME newsletter [13], moderates the Community Memory mailing list [14] for CPSR. Its intention is to document the early history of the computer era in the words of the people who made the history. Recently Robert H. Zakon <zakon at hobbes dot mitre dot org> sent the list a pointer to his excellent Internet timeline -- see [15].

[13] <http://www.reach.com/matrix/meme-archive.html>
[14] <http://www.reach.com/matrix/community-memory.html>
[15] <http://info.isoc.org/guest/zakon/Internet/History/HIT.html>

It's a good thing I'm not aiming to make a living from TBTF anytime soon, according to this opinion from Forrester Research of Cambridge, MA. $3.9 million -- mere pocket change.

>> From Edupage (1996-07-04):

> Most companies banking on selling their content over the Web won't see
> a profit until the year 2000, predicts Forrester Research, which says
> the typical site, such as an electronic newsletter or magazine, will
> lose $3.9 million beyond the initial investment before they start
> making money. "Content providers who joined the Web gold rush find
> themselves tumbling down a long, dark mine shaft. It will be at least
> four years before they see a return on their investments," says the
> report's author. (Investor's Business Daily 3 Jul 96 A5)

>>From Computer Industry Daily (1996-06-25):

> MCI will drop another $60 million to quadruple its Internet
> backbone speed. Throughput will be boosted to 622 Mbps from the
> current 155 Mbps. These enhancements are the second phase of
> MCI's Internet 2000 strategy. Need for the upgrade was signalled
> by a 5,600% increase in Internet traffic since 1994.

In cryptographic circles the common wisdom says that the best protection flows from a system designed around "something you know, something you have, and something you are." Something you know: a password or PIN. Something you have: a personalized smartcard or a torn playing card. Something you are: your face, signature, or fingerprint.

Mytec [16] is a Toronto company that has garnered attention recently for its patented fingerprint-based encryption system based on optical computing. Information from a fingerprint is transformed into a Bioscrypt, or biometric signature encryption. A Bioscrypt is the two-dimensional image of a set of characters (representing an encryption key or a PIN) that has been encrypted by the two-dimensional information in a fingerprint pattern. A Bioscrypt bears no resemblance to the original fingerprint and cannot be reconverted to it. The Bioscrypt cannot be generated or verified from a fingerprint or from anything other than the living person's finger, because variables such as ridge depth, skin elasticity, and moisture content affect the optical calculation whose result is the Bioscrypt.

For an entry point into the subject of biometric encryption, see the Biometric Consortium's site [17]. (Be sure to read the conditions, restrictions, and disclaimers at [18].) The site has links [19] to research or commercial systems for biometric encryption based on: the face -- 5; fingerprint -- 9; hand -- 1; voice -- 8; and signature -- 2.

Some discussion of the Mytec system appeared on Computer Privacy Digest [20] V8 #044 [21]; I read it on on Phil Agre's Red Rock Eater News Service.

[16] <http://www.mytec.com/>
[17] <http://www.vitro.bloomington.in.us:8080/~BC/>
[18] <http://www.nwscc.sea06.navy.mil/warning.html>
[19] <http://www.vitro.bloomington.in.us:8080/~BC/#systems>
[20] <http://www.cs.uwm.edu/faculty/levine/cpd.archive.html>
[21] <http://itu.rdg.ac.uk/misc/Mailing_Lists/cpd/00000023.htm>

Norfolk Island [22] [23], a 4-mile-square dot off the east coast of Australia, was home in the 19th century to some of the descendents of the Bounty mutineers. The island carries on its fine traditions today; an ISP [24] in New South Wales stands to make a lot of money from Mutiny on the Internet. RealNet Access is selling .nf domain names.

The U.S.-based name registry, InterNIC, charges $50 per year and gives out domain names on a first-come, first-served basis. The NIC has elaborate procedures for attempting to resolve trademark/domain name disputes under essentially hopeless circumstances: many identical trademarks can coexist per country in disparate industries, whereas domain namespace is unified across the world. The arbitration rules favor trademark holders at the expense of domain-name holders. The result is grief and lawsuits [25] [26] [27] [28].

All right, so you can't have mcdonalds.com. How about mcdonalds.com.nf? It's yours for $250 and $50 per year. Or you could opt for the greater elegance and concision of mcdonalds.nf: $1,000 and $100 per year. At this writing these names have not been claimed.

The business would seem to be aboveboard, assuming that RealNet Access has the legitimate right to subdivide the .nf domain. Upon successful purchase they promise to grant you title to your .nf domain name and to forward a certificate of ownership. How you deal with McDonalds Corporation's lawyers is up to you.

Information on the Norfolk Island namespace landrush was first posted to Usenet by Marshall Clow <mclow at mailhost2 dot csusm dot edu>.

[22] <http://www.anbg.gov.au/norfolk.gardens/norfolk.loc.gif>
[23] <http://gwb.com.au/gwb/nfk/nfkmain.html>
[24] <http://www.names.nf/>
[25] <http://infolawalert.com/stories/051796b.html>
[26] <http://infolawalert.com/source/src061496_dc_intro.html>
[27] <http://infolawalert.com/stories/061496b.html>
[28] <http://infolawalert.com/articles/960628net_sol_interpleader.html>