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

TBTF for 1996-10-31: Information super chariot track

Keith Dawson (dawson at world dot std dot com)
Fri, 1 Nov 1996 00:00:34 -0400


Backbone woes and the end of flat-rate pricing

Hal Varian, dean of the School of Information and Management Systems at U.C. Berkeley, has for some time been predicting the traffic woes and backbone clogging we now see increasingly on the Internet. He recommends that as the current no-cost, trust model of data exchange breaks down, the top-level backbone carriers institute usage-based pricing: users paying more to get higher levels of service guarantees. See [1] for a summary of a talk Varian gave last spring on this topic. For a lucid background picture of the way the network is put together now I recommend this essay [2] on the New York Times site (free registration is required) -- the authors claim that there is no economic model to support the continued growth of high-bandwidth use on the Internet. Today's TechKnow Times tantalizes with the news (but no details) of a major carrier having removed one of the regional carriers from its routing tables, rendering some U.S. east-coast sites unreachable for the last five days. And Edupage recently quoted MCI's Vint Cerf, one of the Internet's inventors, sounding like a born-again Varianista. (MCI's wires carry 90% of the Internet's backbone traffic.)

>>From Edupage (1996-10-29):

> "Ultimately, it must be the case that the pricing covers the cost of
> service -- and in circumstances where usage is not stable but growing,
> one needs to price according to use. While I am a strong advocate of
> flat-rate methods to stimulate market development, I think one has to
> be realistic about relating price to cost."

[1] <http://www.essential.org/listproc/tap-info/0075.html>
[2] <http://www.nytimes.com/library/cyber/week/0629backbone-money.html>


Flogging key recovery abroad

The success or failure of the U.S. approach to the spread of strong cryptography -- called key recovery, the latest form of key escrow -- will hinge on the Clinton administration's ability to convince other countries to go along. The Netizen last week broke the news [3] that the President is about to appoint a special ambassador whose job will be to apply pressure to U.S. allies and trading partners, getting them to cooperate on the widespread adoption of key-escrow cryptography. This move should surprise no-one who has been paying attention to the debate; FBI Director Louis Freeh's remarks before Congress last June [4] stressed the importance of international cooperation on key escrow. In my opinion the next Administration move to watch for will be the proposal of laws to limit the _import_ of cryptography technology. (Declan McCullogh wrote in a recent dispatch that a government official had hinted as much to him.) Perhaps coincidentally, Japan recently tightened controls on crypto exports [5]; a spokesman claimed that the change was not the result of any U.S. pressure.

[3] <http://www.netizen.com/netizen/96/43/index3a.html>
[4] <http://www.crypto.com/events/072596/freeh.html>
[5] <http://pathfinder dot com/@@VVZl0QUAbzzTmyvI/Netly/daily/961029 dot html>


A plan for adding new domains hits a speed bump

Followup: TBTF for 1996-08-25 [6]

The proposal [7] originally floated by the Internet Assigned Numbers Authority for adding up to 150 more top-level domains has been sidelined by one of the IANA's sponsoring organizations, the Internet Society. The proposal caused "considerable international debate on various aspects... with no consensus," according to the ISOC's CEO [8]. The matter will be referred to a nine-member international panel for resolution.

[6] <http://www.tbtf.com/archive/1996-08-25.html>
[7] <http://www.iana.org/iana/registries.html>
[8] <http://www.isoc.org/whatsnew/iahc.html>


Cyber Promotions, the largest email spammer, staggers

News for students of Internet business models: one that doesn't work is sending hundreds of thousands of unsolicited email messages a day, using forged return addresses, to online-service subscribers. Things have not been breaking well for Cyber Promotions. Days after a court prevented AOL from blocking Cyber Promo's email from its customers [9], the online service won a reversal of that ruling [11]. Soon AOL had launched its eagerly awaited Preferred Mail service that lets its customers individually choose whether to block such mail; Seidman's Online Insider for 10/25 (currently at [12], soon at [13]) lists 50 sites on AOL's initial default blocking list. Cyber Promo lost a court fight with Concentric Networks [10] and was enjoined from forging return addresses from that ISP. On 10/24 Compuserve won a similar injunction [14]. (This note [15] posted to the e$ mailing list demonstrates that on 10/18 a Cyber Promo machine was indeed manifesting a Compuserve IP address.) A week later the company settled a Prodigy lawsuit by agreeing to abide by similar restrictions. On 10/25 an article on ZDnet [16] claimed that the company, having been cut off by its latest ISP, SprintNet, had suspended service and was near bankruptcy.

[9] <http://www.tbtf.com/archive/1996-09-08.html>
[10] <http://www.tbtf.com/archive/1996-10-09.html>
[11] <http://www-db.aol.com/corp/news/press/view?release=53&>
[12] <http://techweb.cmp.com/ng/online_insider/current.htm>
[13] <http://techweb.cmp.com/ng/online_insider/102596.htm>
[14] <http://world.compuserve.com/world/news/order.html>
[15] <http://www.tbtf.com/resource/cyberpro-cs.txt>
[16] <http://www.zdnet.com/intweek/daily/961024x.html>


Trustmarks for electronic transactions

The Electronic Frontier Foundation is working with CommerceNet on a new initiative to bolster trust and confidence in electronic transactions. eTRUST [17] will establish and evolve guidelines on issues of privacy, security, and authentication. Their first project addresses online privacy. The organization has developed and will license to online merchants recognizable and credible "trustmarks" [18] of privacy and security. eTRUST plans to run a pilot program later this year and to announce a commercial offering early next year.

[17] <http://www.etrust.org/>
[18] <http://www.etrust.org/05eff.html>


Oracle and Netscape team up (against guess who)

The two companies revealed plans [19] to bundle each other's software and to work together on a network computer that runs on an Intel chip. The NC will officially be introduced at Oracle's trade show on 11/4, but the company has been showing prototypes for a while now. The newfound cooperation puts an end, for now, to the two companies' recent public feuding.

[19] <http://www.news.com/News/Item/0%2C4%2C4838%2C00.html?nd>


Halloween special: Web crawlers

Do you know how to bring your Web site's robot visitors to heel? The polite ones knock before entering, and the protocol for knocking is spelled out in the venerable Robot Exclusion Standard [20]. Its author, Martijn Koster <m.koster at webcrawler dot com>, also maintains a list of known robots [21] that currently details 111 of the beasties. Finally, to learn more than you ever wanted to know about robots, worms, spiders, crawlers, and other assorted Web critters visit the new BotSpot [22].

[20] <http://info.webcrawler.com/mak/projects/robots/norobots.html>
[21] <http://info.webcrawler.com/mak/projects/robots/active.html>
[22] <http://www.botspot.com/>


Science Surf

William Calvin must be one energetic guy: he's a neurophysiologist at the University of Washington, the author of nine books, and the creator at irregular intervals of Science Surf [23], a must-visit destination for anyone with an interest in science, good writing, or Web design. (Use Internet Explorer >= 2.1 or Netscape 3.0: no funny plug-ins required, but you'll be glad of the support for color in table cells.) I've got a URL-minder [24] on this one.

[23] <http://www.well.com/~wcalvin/scisurf/scisurf.html>
[24] <http://www.netmind.com/URL-minder/>


Information super chariot track

On September 5 of this year I was fortunate enough to catch a broadcast on National Public Radio of a piece by Dan Charles titled "Some Say the Internet Will Go the Same Route as Early Networks." I have a transcript -- copyright forbids posting it, but I will outline it here for you. The thesis is that each of the successive waves of information technology to transform American society has followed in the footsteps (or more literally, in the trackbeds) of the railroads. Railroad historian John Hankey is quoted:

"Almost every claim that the folks putting together the new Information Age make was true of the 1830s with the railroad -- the claims that it would bring us closer together as a people, or that it would revolutionize the way we bought and sold things, or that it would change our culture."

In the trackbed of the Baltimore & Ohio railroad, the nation's oldest, Samuel F. B. Morse buried lead-covered copper cable to carry the first telegraphic messages. Later, high-tension electrical wires followed the same well-worn path. Today two bundles of optical fibers, one laid down by MCI and the other by Sprint, carry rivers of data. (Recall that Sprint grew out of the Southern Pacific Railroad -- that's where the "Spr" comes from.)

The analogy goes deeper. The earlier information technologies evolved in similar ways, and a prudent observer might bet that the Internet will yet again follow in their tracks. A brief period of frantic growth, with dozens of companies building their own independent networks and standards, will be followed by a rapid consolidation into the hands of the few big companies that own the wires -- MCI, Sprint, and AT&T. And possibly WorldNet, the nation's fourth-largest phone company, whose long-distance business is based on the natural-gas pipelines in which they laid optical fiber when their energy business began to decline. Right of way is still king. There are rumors that the railroads are quietly laying down their own fiber lines, angling to get a piece of this latest network business; ditto the electric power companies.

We could carry the historical analogy further. This hoary Internet fable [25] contends that the reason our railroads use a gauge of 4 feet, 8-1/2 inches is that they were built by British railroaders, whose rails were spaced equal to the earlier wagon ruts, which date back to Roman roads, which were rutted by war chariots. Mil specs live forever.

Note added 1996-11-01: Thanks to E. M. Ganin <eldad at hi dot com>, who succeded in the quest I gave up on -- he found the original author of How Mil Specs Live Forever [25]: it's Bill Innanen <wgi at aplcomm dot jhuapl dot edu>. Here is Bill's elaboration on the creation of this gem.
I wrote the "MilSpec" story on 9 Feb 1994 as a humor posting for a private mailing list. You can see the original posting on my web site at [25]. I had never seen the story in writing, but I had heard a similar piece on the radio some years previous. I think it was on the CBS radio series "The Osgood Files" but I'm very uncertain of that. It'd be something that Charles Osgood would do, though. Because I deal professionally with the military R&D world, the story stuck with me through the years. Poking fun at MilSpecs really tickled my funny bone.

I did a little research on the subject prior to writing up the story. I left out a few of the details that I had learned, like the fact that the earliest known use of that exact gauge was in a particular coal mine in Northumberland, and that the ruts in the Roman roads would probably have accommodated wheel gauges that varied by a few inches. Leaving out some of those details made for a far better story without actually making it fictitious. There were clearly many factors that went into the adoption of the US standard railroad guage. This was just one of them -- the funniest one. The post was never meant to be a serious historical thesis -- just a good chuckle. Enjoy.

One embellishment on the tale traces it back to Abraham Lincoln who supposedly discovered the tortuous track back to the Roman chariots. Another asserts that Napoleon's Russian defeat by Generals January and February can be laid to the difficulty his army had in negotiating the east European wagon ruts of non-Roman gauge.

We'll leave the last word to Matt Wall <wall at cmu dot edu>, who posted a definitive rebuttal [26] to the pretty fable of Roman origins for the modern rail. The rebuttal does, however, buttress the predictive arguments of Dan Charles cited above. Expect infrastructure consolidation.

[25] <http://www.clark.net/pub/innanen/wgi_postings/wgi_val-l_940209.txt>
[26] <http://www.tbtf.com/resource/rail-debunk.txt>


>>TBTF was reviewed, favorably I'm happy to report, in the October 21 Netsurfer Digest <http://www.netsurf.com/nsd/nsd.96.10.21.html>. In the ensuing week this publication's direct subscriber base rose by 25% to break 2,000. If any of you don't know about the excellent Netsurfer family of Web resources, by all means visit <http://www.netsurf.com/> now.


>>E.Commerce Today -- this commercial publication provided background information for some of the pieces in this issue of TBTF. For complete subscription information see <http://www.tbtf.com/resource/e.commerce-today.txt>.

>>TechKnow Times -- mail techknow-request@techknowtimes.com with subject: subscribe . Web site at <http://www.TechKnowTimes.com/>.

>>Edupage -- mail listproc@educom.unc.edu without subject and with message: subscribe edupage Your Name . Web site at <http://www.educom.edu/>.

>>Seidman's Online Insider -- mail listserv@peach.ease.lsoft.com with message: subscribe online-l Your Name . Web site at <http://techweb.cmp.com/ng/online_insider/>.

>>e$ -- mail majordomo@thumper.vmeng.com without subject and with message: subscribe e$ . Web site at <http://www.vmeng.com/rah/>.

TBTF alerts you weekly to bellwethers in computer and communications tech-
nology, with special attention to commerce on the Internet. See the ar-
chive at <http://www.tbtf.com/>. To subscribe send the message "subscribe"
to tbtf-request@world.std.com. TBTF is Copyright 1996 by Keith Dawson,
<dawson at world dot std dot com>. Commercial use prohibited. For non-commercial
purposes please forward and post as you see fit.
Keith Dawson    dawson at world dot std dot com
Layer of ash separates morning and evening milk.