A new patent reveals how they plan to skirt Intel's patents

For over a year folks have been reading the tea leaves of Transmeta patent filings [1] to divine what the secretive company is in business for. Why do we care? Mostly because Microsoft co-founder Paul Allen is an investor and Linus Torvalds, father of Linux, works there [2]. Transmeta's new patent [3] reveals, to those who troll these deep waters, that the company is developing a processor capable of running the Intel instruction set (no surprise so far) while skating around Intel's own technology patents. This CNet story [4] notes that Torvalds has hinted that Transmeta might debut its products at the Comdex trade show in November [5]. And a TBTF informant who must remain nameless (heck, I don't know who s/he is) claims that Microsoft's Windows 2000 kernel / driver team possesses detailed knowledge of Transmeta's strategy and operates under a non-disclosure agreement with the company. Who knows, in a few years we may speak of the Winsmeta duopoly.

[1] http://tbtf.com/archive/1998-11-17.html#s05
[2] http://tbtf.com/archive/1997-03-09.html#s07
[3] http://164.195.100.11/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=...
[4] http://www.news.com/News/Item/Textonly/0,25,424001,00.html
[5] http://www.news.com/News/Item/Textonly/0,25,122755,00.html

The three parties have been wrangling over contractual terms for the last year. Last week they announced a complex series of agreements that resolve all of the issues outstanding among them, including funding for ICANN's continuing operations. The best summary I've found of the interlocking agreements is this fact sheet [6] on Commerce's site. The agreements could come into effect as early as November, after ICANN takes public comment and ratifies them.

Highlights:

• NSI assents to ICANN's authority and agrees to sign a modified Registrar Agreement.

• Commerce takes over operation of the InterNIC.

• The fee NSI charges to competitive registrars drops from $9 to $6.

• NSI agrees in principle to a per-name fee to fund ICANN's operations, provided that NSI does not owe more than $2M under such a program. NSI hands over $1.5M to ICANN immediately.

• NSI continues to run the authoritative root server for at least four years. Even after its eventual transfer to ICANN, Commerce continues to assert policy authority to direct this server. (I wonder what the EU thinks of this provision.)

• NSI must totally separate its registry and registrar functions. If it accomplishes this within 18 months then it can hold onto the root server for an additional four years.

• NSI effectively gives up the claim that it owns the intellectual property represented by the .com/.org/.net database.

With the contract fight behind them, ICANN moved forward with their proposal for a uniform policy for resolving disputes over domain names [7], [8]. Its main goals are to render domain-name hoarding profitless and to remove most disputes from the courts in favor of binding arbitration. ICANN will take public comments [8] on the proposal until 13 October.

At the recent conference [9] "Governing the Commons: The Future of Global Internet Administration," many participants were critical of ICANN's attempts to establish Internet policy, according to this account [10] written by Ted Byfield <tbyfield at panix dot com> for the German magazine Telepolis. Byfield notes that ICANN has blown past the already controversial proposals of the IAHC-gTLD-MoU-CORE group [11], [12], which wanted to establish equitable dispute resolution mechanisms. ICANN proposes a much stronger uniform dispute resolution policy, drawing even more fire.

[6] http://www.ntia.doc.gov/ntiahome/domainname/agreements/summary-factsheet.htm
[7] http://www.news.com/News/Item/Textonly/0,25,805704,00.html
[8] http://www.icann.org/udrp/udrp.htm
[9] http://www.cpsr.org/conferences/dns99/dnsconf99.htm
[10] http://www.heise.de/tp/english/inhalt/te/5345/1.html
[11] http://www.gtld-mou.org/
[12] http://tbtf.com/resource/domain-name-hist.html

Declare defeat, but stay in

On 16 September the administration announced changes in the US cryptography export regime. Like numerous other changes in the past, this one was presented as a relaxation of the rules that will benefit consumers. It's far from clear that this is the case.

Once the new rules go into effect in December, after a one-time review any retail product featuring encryption of any strength will be exportable to individuals and companies -- but not to governments -- in all but 7 countries worldwide. This relaxation is tied to funding for a new FBI research lab and to a disturbing loosening of the rules of evidence in court cases that involve encryption.

The Electronic Privacy Information Center links the White House announcement, commentary, and analysis from this page [13]. EPIC remains agnostic on the proposals. General counsel David Sobel said, "It appears that the FBI and large computer companies have reached an agreement on encryption, but that is not necessarily in the interest of the average computer user."

The legislative vehicle for these new initiatives is the selfsame Cyberspace Electronic Security Act that, in an earlier draft, would have allowed secret police break-ins to alter computer equipment [14]. That provision is gone now; it was probably a trial balloon anyway.

A week after the latest proposals were announced, EPIC's Mark Rotenberg found himself sharing a conference panel with William Reinsch, the administration official tasked with carrying out US crypto export policy. Rotenberg later described his address to the politech mailing list:

International group breaks the seventh Certicom challenge

Irish mathematician Robert Harley announced [15] that his team has cracked the seventh and most difficult Certicom ECC Challenge problem to date. Certicom has confirmed the correct result [16]. So far seven Certicomm exercises and challenges have been cracked since December 1997; Harley's growing team has broken each one of them. The solution required 16,000 MIPS-years -- twice the effort of the recently broken, 512-bit RSA-155 [17]. The team struck it lucky, finding the solution in less than a third of the expected time. The distributed computation was run by 195 volunteers, on a total of 740 computers, over 40 days.

While this result strengthens the case of those who contend on theoretical grounds that a crypto key based on ECDL (Elliptic Curve Discrete Logarithms) is inherently harder to break than an RSA key, it does not prove that assertion. Rather, it indicates that at the current state of the art, the best mathematical tools and algorithms known for cracking ECDL take longer to run than the best tools known for cracking RSA.

On 12 September I posted as a Tasty Bit of the Day Harley's call for more machines to throw at the problem; others, including TechDirt, publicized it as well. This graph [18], adapted from Harley's site, rather dramatically shows the effect of the call for participants.

[15] http://cristal.inria.fr/~harley/ecdl/
[16] http://www.certicom.com/press/99/sept2899.htm
[17] http://tbtf.com/archive/1999-08-30.html#s01
[18] http://tbtf.com/pics/ecdl-prog.gif

British Sunday paper gets trolled

This story in The Times (UK) [19] claims that a European Institute of Quantum Computing Network has been hastily formed to develop commercial banking codes based on quantum entanglement. The newspaper claims:

New free service should prove a boon to list managers, members, and those in need of ad hoc groupware

Last week Internicity, Inc. released Take It Offline [21]. TBTF is proud to offer you this exclusive first look. (Full disclosure: Steve Yost, Internicity's principal, is a TBTF Irregular [22] and a friend of mine. I offered him ideas and advice from the earliest days of Take It Offline, and the TBTF Irregulars supplied beta feedback. I don't have any financial interest in Internicity.)

TIO provides a convenient, lightweight venue for ad-hoc, online group discussions. Did someone on your mailing list just raise an off-topic but intriguing idea? In less than a minute you can create a private TIO discussion space and post its URL to your list. Then anyone interested in following the diversionary thread can participate at Take It Offline. The mailing list stays focused. Once the TIO discussion winds down, the thread stays live, so you get no 404s from a mailing list's Web archive or from search spiders.

TIO can be useful in the absence of a mailing list. Say you need to coordinate a seminar schedule involving 10 people. You can create a TIO space and mail its URL to the 10 individuals, then work out the details in Take It Offline. Those who wish to can get email each time a note is added to the discussion; a daily digest is also available.

The site is lean and speedy, light on graphics, clean and attractive. The privacy policy is featured prominently and it is aggressively visitor-friendly. Posters to TIO can use any name they like; no registration or passwords are required. An email address is needed only to start a thread. Cookies are used only for visitor convenience; the site works fine if you refuse them.

In deference to the recent Jargon Scout entry [23], Internicity has also registered the name Takeitofflist.com.

Take It Offline can provide a free forum for mailing lists, such as this one, that don't offer threaded discussions. Let's try it now. Visit this TIO space [24] if you want to explore TIO's implications for the dynamics of mailing lists or the workings of hypertexts. I'll be following this thread closely and posting to it from time to time.

[21] http://www.quicktopic.com/
[22] http://tbtf.com/the-irregulars.html
[23] http://tbtf.com/jargon-scout.html#e2e
[24] http://www.quicktopic.com/1/H/emxb7z9ubBnT5eAuc8l.html

But who can replace a man?

Researchers at the University of Southern California announced [25] a neural network system, curiously unnamed, that they claim performs better than humans at recognizing words under noisy conditions. In the tests described, the USC system was pitted against human subjects in the task of picking out individual words amid varying amounts of white noise or conversational babble. The system performed as well or better than the human subjects across the board; the noisier the conditions the greater its advantage [26].

The researchers say that this performance, far beyond that of conventional voice-recognition systems, stems from the unique neuron-mimicking chips they have developed. Like neurons, the chips signal by varying their rate of output. Previous neural circuits kept their output clocked, ignoring this timing aspect of the way biological neurons operate.

It's unclear how well such a system would scale. The reported experiments used only four separate words, on which the USC system had been trained. Adding more words might dilute its accuracy; such has been the experience of other neural networks. But USC obtained their results with a circuit of only 33 neuromime chips, versus the hundreds or thousands of (software or hardware) simulated neurons used in other research.

I can't judge the significance of this announcement; I've seen some skepticism directed towards it but no substantial arguments. Perhaps the researchers are onto something truly important with their unclocked neuromimes.

[25] http://www.sciencedaily.com/releases/1999/10/991001064257.htm
[26] http://tbtf.com/pics/USC-neural.gif

A little maze of twisty items, all different

Amazon's zShops

Everyone writes about Amazon.com. That's because almost anything they do ends up looking like a leading indicator for where Internet commerce is headed next. Amazon's latest move is a stunner, but the jury is definitely out on whether or not it's a good idea. Amazon has introduced zShops [27], [28], a way for a small business or an individual to offer anything for sale to Amazon's millions of daily visitors, using Amazon's fabled One Click Ordering. On the one hand, Amazon continues to do what they've always done best: leave the competition at the last turn scratching their heads. On the other hand, Amazon looks set to squander its hard-won brand name by representing hundreds of thousands of items and merchants that don't measure up to its quality standards.

[27] http://www.seattle-pi.com/business/amaz01.shtml
[28] http://auctions.amazon.com/exec/varzea/subst/home/fixed.html

Google site comes out of beta

Jochen Schwarze <jochen dot schwarze at orthogon dot de> was the first to send word of the formal launch of the Google search site [29]. The company introduces a feature called GoogleScout, which seems to be a form of "more like this link." TBTF profiled Google on 1998-05-11 [30] -- the first press coverage for the site in English, before its founders had left Stanford.

[29] http://www.google.com/pressrel/pressrelease4.html
[30] http://tbtf.com/archive/1998-05-11.html#s08

Eclipse followups

The article "The partial eclipse at the Duomo" in the previous issue [31] prompted these notable eclipse pointers.

• Mark Dionne <mdionne at mediaone dot net> directs our attention to this stunning photograph [32], taken from the Mir space station, of the August 11 total eclipse on the face of Europe. (Two weeks later the last Mir crew turned out the lights and pulled the plug [33].)

• Peter Kaiser <kaiser at acm dot org> kindly gave permission to post on the TBTF archive his account [34] of travelling to view the eclipse. It's not what you might be expecting. Kaiser gives step-by-step instructions for recreating his eclipse experience in New York City. On the west side. On Riverside Drive. Oh, just go read it [34].

[31] http://tbtf.com/archive/1999-09-11.html#s08
[32] http://antwrp.gsfc.nasa.gov/apod/ap990830.html
[33] http://www.cnn.com/TECH/space/9908/28/mir.farewell.reut/index.html
[34] http://tbtf.com/resource/kaiser-eclipse.html

Think Open Source guarantees you can know what a program does? Think again

This classic paper [35] by Ken Thompson, co-inventor of Unix, is disquieting in the extreme. It is Thompson's 1984 acceptance speech for the ACM's Turing Award. Understanding it requires some grasp of the mechanics of programming. That said, those who read and grok Reflections on Trusting Trust will emerge considerably more paranoid than they went in. The effect is likely to be permanent.

[35] http://www.acm.org/classics/sep95/
[36] http://www.quicktopic.com/1/H/ryToV91WMVflxRhI63YY.html

What do you call an Easter egg in a search engine?

Easter eggs, as usually defined, are the amusing personal messages that programmers leave buried in commercial software [37]. The Web is broadening the possibilities to which an Easter egg can aspire. The first search-engine egg I've encountered is built into Google's priority rules. Whose home page do you suppose tops Google's reply when you enter more evil than satan himself [38]? This offbeat discovery has been circulating on various mailing lists in recent days and was picked up by the Memepool [39] blog [40]. If you've seen any other search eggs, or noted other directions in which Easter eggs are expanding, let's take it offlist here [41].

  "more evil"                         Microsoft is match 1
  "more products"                     Microsoft is match 2, Adobe 1
  "more evil than gates himself"      Microsoft is match 1
  "more good than satan himself"      Microsoft is match 1
  "not as evil as satan himself"      Microsoft is match 1
  "more evil than your mom"           Microsoft is match 1
  "more pretty than your girlfriend"  Yahoo is match 1

[37] http://tbtf.com/archive/1997-01-29.html#s10
[38] http://www.google.com/search?q=more+evil+than+satan+himself
[39] http://www.memepool.com/
[40] http://tbtf.com/jargon-scout.html#blog
[41] http://www.quicktopic.com/1/H/WuyYlLViKRbpQsFQnC.html

• The Take It Offline service and its impact on the dynamics of email lists and discussions:
  http://www.quicktopic.com/1/H/RsNRIBn9EJCQYhlafHO.html

• Ken Thompson's paper Reflections on Trusting Trust:
  http://www.quicktopic.com/1/H/ryToV91WMVflxRhI63YY.html

• Easter eggs built into search-engine rules:
  http://www.quicktopic.com/1/H/WuyYlLViKRbpQsFQnC.html

• Other comments about this issue of TBTF:
  http://www.quicktopic.com/1/H/RsNRIBn9EJCQYhlafHO.html

TBTF home and archive at http://tbtf.com/ . To (un)subscribe send
the message "(un)subscribe" to tbtf-request@tbtf.com. TBTF is Copy-
right 1994-1999 by Keith Dawson, <dawson dot tbtf at gmail dot com>. Commercial
use prohibited. For non-commercial purposes please forward, post,
and link as you see fit.
_______________________________________________
Keith Dawson    dawson dot tbtf at gmail dot com
Layer of ash separates morning and evening milk.