(A Javascript-enabled browser is required to email me.)
TBTF logo

TBTF for 1999-10-05: Offlist

Keith Dawson (dawson dot tbtf at gmail dot com)
Tue, 5 Oct 18:04:14 -0400

C o n t e n t s

Transmeta aims squarely at Intel

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
[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


Threads Domain name policy
See also TBTF for
2000-04-19, 03-31, 1999-12-16, 10-05, 08-30, 08-16, 07-26, 07-19, 07-08, 06-14, 05-22, more...

ICANN, NSI, Commerce kiss and make up

An interlocking set of agreements removes some roadblocks stalling domain-name reform

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.


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


Threads Cryptography export policy
See also TBTF for
2000-02-06, 1999-10-05, 08-30, 08-23, 08-16, 07-26, 05-22, 05-08, 04-21, 03-01, 01-26, more...

US cryptography export rules eased

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:

I opened by quoting Senator Aiken's line regarding Vietnam that the US should "declare victory and then get out." I suggested that with the crypto issue, the Administration has decided to "declare defeat, but stay in."
[13] http://www.epic.org/crypto/announce_9_16.html
[14] http://tbtf.com/archive/1999-08-23.html#s01

space ______
Threads Using the Internet as a massively parallel computer
See also TBTF for
2000-03-31, 1999-08-30, 07-19, 01-26, 1998-03-02, 1997-10-27, 09-08, 09-01, 06-23, 01-29, 1996-12-02

97 >> 512

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.

Certicom Challenge machines 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

space ______

Threads Quantum computers and quantum physics
See also TBTF for
1999-10-05, 01-04, 1998-11-03, 10-27, 09-14, 03-09, 03-02, 02-23, 1997-11-24, 09-15, 05-22, more...

A working Israeli quantum computer (not)

British Sunday paper gets trolled

This story in emended 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:

The institute was founded a few weeks after news leaked from the Israel's Weizmann Institute that it was using a mixture of quantum computing and special optical technology to break the RSA-512 code, the system used by the European banking system. It claims it has developed a hand-held device that can break the code in 12 microseconds.
The "special optical technology" sure sounds like Shamir's TWINKLE [20]. An opto-electronic sieving device, which as far as I know has never been constructed, is exactly what you'd want if your goal was to accelerate a brute-force attack on RSA-512. But 12 microseconds? It seems unlikely in the extreme. Its inventor estimates that TWINKLE would speed up sieving by a factor of 1000 -- that is, for RSA-512, this step would take hours instead of months. And quantum computers are generally thought to be years from practical realization, if not decades. Here's the succinct dismissal of crypto expert Peter Gutmann <pgut001 at cs dot auckland dot ac dot nz>, writing on the EUcrypto mailing list:
I would say the quantum crypto aspect is at least as accurate as the confused gobbledigook in the rest of the article, which looks like it was cobbled together from pieces of reports on TWINKLE, the factoring of RSA-512 in August, and a sales pitch for some crowd in Europe. I assume the latter was the driving force behind the story.
[19] http://www.sunday-times.co.uk/news/pages/tim/99/09/29/timintint02001.html?1341861
[20] http://tbtf.com/archive/1999-05-08.html#s02


Take It Offline: enhancing online discussion

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


Neural network said to beat people at speech recognition

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


Quick bits and followups

A little maze of twisty items, all different

bul 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

bul 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

bul Eclipse followups

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

[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


On trusting code

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.

The moral is obvious. You can't trust code that you did not totally create yourself. (Especially code from companies that employ people like me.) No amount of source-level verification or scrutiny will protect you from using untrusted code... As the level of program gets lower, these [Trojans] will be harder and harder to detect. A well installed microcode bug will be almost impossible to detect.
I'm interested in what non-programmers are able to make of Thompson's revelations; let's take it offline [36]. Other insights on this paper are welcome too.

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


Egg hunt

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].

Note added 1999-10-06: Joshua Eli Schachter <joshua at memepool dot com>, proprietor of Memepool and a TBTF Irregular, writes:
Ugh. I hate the terms "blog" and "weblog" in general. Especially since those are usually the clicktrails of a single person, and memepool is the combined efforts of ~70 people. But thanks for the link :)
Note added 1999-10-06: Posters to the TIO discussion area, led by TBTF Irregular Danny O'Brien, have pointed out that Microsoft's home page comes up at the top of many Google searches. The engine ranks pages based on others' links to them, and many people link to Microsoft. Some fraction of these links happen to be attached to text containing the words "evil" and/or "satan."

Concensus: it's not an Easter egg, it's behavior as designed.

Michael Moncur added this thought:

[T]his is a misinformed lawsuit waiting to happen. Just wait until some commission-hungry lawyer discovers that typing "bad customer service" returns Symantec's customer service page. "Worst computer" returns a link to Compaq. "Worst operating system" and "worst browser" link to Microsoft (of course). "Worst bookstore" links to Amazon.com (as does "best bookstore".)
Note added 1999-10-07: David Helder <dhelder at umich dot edu> posted this illuminating enumeration to a private mailing list; it is quoted here by permission.
Here's some other strings I tried and what I got:
  "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
Also, "murder can be fun" now yields 1) ReSearch on Zines, 2) [the] "Murder can be fun" zine site, and 3) Amazon.com. About a year ago... it yielded Disney.

[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

O f f l i s t    D i s c u s s i o n s

bul These are the forums I've set up at Take It Offline for those who want to comment on and discuss this issue's articles. I'll be monitoring and posting to these forums actively until at least 15 October.

N o t e s

bul Four days on a windjammer off the coast of Maine was the longest I've been out of reach of IP tone since 1995; couldn't even raise a cell-phone response. A few of you were kind enough to inquire about the fate of a certain schooner in the winds of Hurricane Floyd. We truncated the sail a half day early on Thursday morning, fleeing for Camden harbor in rain and 40-mph following winds. Didn't see another single darn fool out on the water all the way home. Nice day for a sail though.

S o u r c e s

bul For a complete list of TBTF's email and Web sources, see http://tbtf.com/sources.html.

B e n e f a c t o r s

bul TBTF is free. If you get value from this publication, please visit the TBTF Benefactors page and consider contributing to its upkeep.

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.



Copyright © 1994-2023 by Keith Dawson. Commercial use prohibited. May be excerpted, mailed, posted, or linked for non-commercial purposes.

Most recently updated 1999-10-09