SMART Letter #43
July 20, 2000

!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*() ------------------------------------------------------------ SMART Letter #43 -- July 20, 2000 Copyright 2000 by David S. Isenberg -- "Other People's Words (OPW)" -- -- 1-888-isen-com ------------------------------------------------------------ !@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*() CONTENTS > Quote of Note: Global Crossing Ad Reveals Continued Content-Conduit Confusion > Smart Remarks from SMART People > Miljenko Horvat on moving to Canada, where the infrastructure is > David P. Reed on McQuillan & Penzias >"Ex-AT&T Believer/Employee" on vultures circling AT&T >"Ed K" on AT&T internal politics > Guy Ward on other dinosaur companies >"Sue Tonym" on AT&T's stumbling basic execution > Jeff Turner on Armstrong's "The sky isn't falling" > Matt Oristano on Armstrong's "The sky isn't falling" > Greg Pelton on Robert Young Pelton > Gary Hughes-Fenchel on war as entertainment > Jay R. Ashworth on Don Tapscott on Ronald Coase > Russell Nelson on Don Tapscott on Ronald Coase > Douglass Turner's view of WAP from Iceland > Alan Schwartz on a Bell Atlantic management problem > Adam Rosenberg on much, MUCH worse phone service > Arthur Einstein's misguided prayer > Peter Varhol makes David I explain independent events > Scott Baker's continuing saga with Bell Atlantic DSL > Peter Cochrane and Ian Dufour get the last word on 'five nines' > Conferences on my Calendar, Copyright Notice, Administrivia ------- QUOTE OF NOTE: from 2-page Global Crossing ad in Wall Street Journal, June 29, 2000. "Global Crossing. It's what happens when the most advanced network on earth meets the world's richest content. . . . A full range of services that meld network and applications." [This is more sad evidence of cable-guy Leo Hindery's muddle-headed mixing of content and conduit at what used to be the world's most promising new telco. Will Gary Winnick wake up before it's too late?] ------- SMART REMARKS FROM SMART PEOPLE: Comments on SMART Letters #38, #39, #40 and #41 --- Miljenko Horvat [] wrote to announce that he is moving from New York to Vancouver, BC, where the infrastructure is (fiber under the building). He says: "the US is not totally losing us since we will keep NewspapersDirect's Park Avenue address but it is certainly true that when you go to Canada and see this semi-socialist infrastructure, you are simply amazed. . . . Even on the residential front, I truly have a choice - I can pick from a few DSL providers, Rogers cable etc." [Note: I am not advocating semi-socialism (down, you jerky knee!). And I *love* the marketplace! But social policy is as given as gravity, and if we don't have good social policy (i.e., that encourages free markets) then we'll get bad social policy made by giants who want their sub-optimal (and sometimes even evil) monopoly protected. -- David I] --- From David P. Reed [] on Sustaining Technology's Last Stand -- SMART Letter #38: "McQuillan is a smart guy - I wish he'd take firmer positions. He always seems to have stuff on the one hand and on the other hand, staying safely above the fray. "Penzias, on the other hand, seems to get it. He understands the fundamental tradeoff of overprovisioning and delay. He sees why optimization of every last ounce of capacity is silly, and why billing in complex ways to micromanage resources that are ever cheaper is silly." --- From "ex-AT&T believer/employee", name withheld by request, on AT&T Circles the Drain -- SMART Letter #39: "The vultures have already arrived to dine on AT&T's carcass. These birds include: ...the whole Internet movement, ...RBOC entry into long distance, ...landline-LD as a dying business model, wireless IP and WAP, ...AOL Instant Messenger (especially v5.0 with the 'talk' feature), ...NextGen IP telcos, ISPs who undercut AT&T's data prices [Yipes! -- David I], "AT&T's bloated, redundant, risk-averse bureaucracy understands it needs to do something soon but still has momentum and hierarchy working against it. "Meanwhile, employee morale spirals down. Vested, long- time (10-15 year) employees are quitting en mass. There are widespread but 'quiet' layoffs (10 people here, 10 people there). Entire departments are put "at risk". Employee benefits are shrinking; medical benefits are much reduced, and legal, vision, prescription, etc. are all gone." --- From "Ed K" on AT&T Circles the Drain -- SMART Letter #39: "I, too, am ex-AT&T. I, too, was loyal to the death star, putting in long hours on projects because I saw how they could help the company in the long run. Then I saw my recommendations ignored until, in one case, the infamous McKinsey consultancy made the same recommendations I made two years earlier -- these, of course, were implemented with dispatch. "Internal political considerations made it too risky for upper management to listen to AT&T employees. Once I saw someone actually stick their neck out against upper management's wisdom, and the risk paid off. Then the same upper management that had previously trashed the idea jumped in to take credit for it. "The rest of AT&T liked to criticize the Labs for figuring out solutions to the 9th significant digit. But no, the problem clearly was inept management." --- From Guy Ward [] on AT&T Circles the Drain -- SMART Letter #39: "The analysis does not just apply to AT&T; other dinosaur companies are in similar trouble. "I just left [a European telecom equipment manufacturer] after 15 years. Companies like [this] are faced with falling margins across a broad range of telecommunications equipment. They try to maintain competitive position by scouring the world to gain access to the new technologies have failed to develop themselves. 'This is not a fun time for funds managers -- 'safe' investments in top companies may be more risky than investments in dot coms." --- From "Sue Tonym" on AT&T Circles the Drain -- SMART Letter #39: "One thought on your painfully accurate column: It's not just the grand strategy, it's the basic execution. I had an atrocious customer service experience with AT&T. As a result I dropped AT&T long distance. "Now AT&T has the local cable monopoly, so I'm a reluctant AT&T customer again. Last week one of the local channels was screwed up, and my [spouse] called AT&T customer service. The quality of the experience resembled that with AT&T long distance. Guess who's switching to a dish and DSL real soon now? "While Armstrong and Nagel play with high strategy, they are letting the fundamentals of operation and execution rot -- keeping your existing customers happy is basic to a subscriber business." --- From Jeff Turner [] in response to the challenge [in SMART Letter #39] to relate C. Michael Armstrong's "The sky has not fallen" quote to "The Innovator's Dilemma": "It's a classic case of alternative, supposedly inferior technologies (IP telephony, email, chat and voice chat, instant messaging, etc.) beginning at the fringe, then moving mainstream. It's only a matter of time before alternative voice technologies begin to appeal to AT&T's most lucrative customers. Then AT&T will be left with a network of buggy whips on a packet autobahn. "Soon there will be no need to pay for two separate connections, one for high-speed data and and one for telephony. Local and LD telephony will become less valuable elements in a bundle, but AT&T won't be able to raise prices on cable modem service because alternative high-speed access will be available. "The final kicker: no way can AT&T get near the cable customer counts they told Wall Street last winter (and reiterated [in early May]. They don't have automated order processing & provisioning yet, and they don't have enough installers to meet demand. (They do have demand, although one can only wonder for how long.) So look for further stock weakness and executive reshuffling after they miss their numbers at the end of the year." ----- From Matt Oristano [] in response to the challenge [in SMART Letter #39] to relate C. Michael Armstrong's "The sky has not fallen" quote to "The Innovator's Dilemma": "I dutifully scanned my dog-eared copy of 'Dilemma', only to find that almost every sentence in the book is entirely and pungently relevant to Mr. Armstrong's quote. I'd really like to know if he's read the book. "Since your contests have entirely elastic rules, I'm going to stretch them by suggesting more than one quote: 1. p xxiii: 'In their efforts to stay ahead by developing competitively superior products, many companies *don't realize the speed* at which they are moving up market, over-satisfying the needs of their original customers. . . . In doing so, they create a vacuum at lower price points into which competitors employing disruptive technologies can enter.' "In other words, Mr. Armstrong, we don't need to hear a pin drop, we don't need to use our 'golden ears'. Your network quality has overshot the needs of the majority of your customers, and is being attacked from below by cheaper and more convenient alternatives. 2. p 84: 'The most vexing managerial aspect of this problem of asymmetry, where the easiest path to profit is up, and the most deadly attacks come from below, is that 'good' management - working harder and smarter and being more visionary - doesn't solve the problem.' "AT&T's managers are facing technologies that most of them thought were toys a few years ago. And they still believe that the key to selling more long distance at the highest price is to have the most exquisitely wonderfully best long distance network in the world. 3. p 110: 'The conclusion that a single organization might simply be incapable of competently pursuing disruptive technology, while remaining competitive in mainstream markets, bothers some 'can-do' managers [who] maintain their competitive intensity in the mainstream, while simultaneously trying to pursue disruptive technology. The evidence is strong that such efforts rarely succeed; position in one market will suffer unless two separate organizations, embedded within the appropriate value networks, pursue their separate customers.' "This is what Armstrong is trying to say he's doing, using "separate" organizations. But AT&T Wireless has disruptive issues of its own borne of ignoring CDMA, and the Net2Phone deal is grasping at straws. If Net2Phone eats AT&T's lunch and erases the main AT&T business of selling long distance, Armstrong will want to BE Net2Phone, not have an off-balance-sheet stake in it. 4. p. 54: ' . . . when the slopes [of improvement curves of old and new technologies] differ, new technologies that are initially performance-competitive only within emerging or commercially remote value networks may migrate into other networks, providing a vehicle for innovators in new networks to attack established ones.' "I was struck by Armstrong's quote about how *surprised* everyone was that these new technologies are taking market share. The steeper upward slope of the disruptive technology always surprises the sustaining manager, who looked at it once a long time ago and wrote it off as irrelevant." --- From Greg Pelton [] on You Can't See *That* on TV -- SMART Letter #40: "Robert Young Pelton is real. He's my cousin. He gets into trouble, gets out and likes it. He's just really curious." [Another SMART Person told me, in person, "Robert Young Pelton is a liar. I've been a reporter. No way. He's lying." Then when I inquired further he wrote, "To be fair, he's more a stretcher and an embellisher (in other words, a bullshitter) than an outright fraud."] --- From Gary Hughes-Fenchel[] on You Can't See *That* on TV -- SMART Letter #40: "Pelton is a bit late with the notion that the future of war is entertainment. I was born in '54; it is my belief that television is largely responsible for the way the Vietnam war was prosecuted (for better or for worse)." [I didn't fall off the pumpkin truck yesterday either, and in my view, we're getting better at it. I'm waiting for the next war, where there's a ground camera looking up at the smart bomb's camera as it falls to earth. To optimize ratings, the multinational multimedia companies will show one view on the network of Nation #1 and the other on the network of Nation #2. The cyber-elite will be able to choose its view, but it'll probably be watching its own favorite channel: the market. -- David I] --- From Jay R. Ashworth [], who comments on my summary of Don Tapscott's summary of Ronald Coase's work in You Can't See *That* on TV -- SMART Letter #40: "Don Tapscott says the major driving factor behind the continued coherence of corporations is that there is less friction in transactions inside a corporation than in those *between* corporations. "I don't think so. Liability, and by extrapolation reputation, is a bigger factor. Consumers are more likely to do business with you if you're General Electric than if you're Joe's Stoves and Caskets because they perceive you to be more substantial." --- From Russell Nelson [], who also comments on my summary of Tapscott's summary of Coase in You Can't See *That* on TV -- SMART Letter #40: "Strictly speaking, Coase won the Nobel Prise for the Coase Theorem. The Coase Theorem recognizes that every externality has two parties, and that sometimes the most efficient solution is for the party creating the externality to pay the other party to change. Moreover, IF *transaction costs are zero* (which they never are in the real world -- it would be like friction vanishing) the free market will always come up with the most efficient solution regardless of the starting state. "That's a sublime theorem. It totally changed the field of economics, and was definitely worthy of a Nobel Prize. Nowadays ordinary mortals like myself can say 'Well duh', but isn't that true of every great invention? "I've been teaching myself economics for four or five years now. The overwhelming conclusion is that you need a stupid network -- that central planning cannot work well. The communication and decision problems kill you. By the time you ship in all the information, make the decision, then ship it out again, all your inputs have changed. "You want to hear my prediction? That QoS will NEVER work as anyone is currently architecting it. QoS is not a protocol problem. It is not a technical problem, even. It is an economic problem, and ONLY an economic solution can fix it. "The key thing to understand is that QoS does not manufacture new network resources -- it merely denies access to some." --- From Douglass Turner [] on WAP: "I spent eight mostly joyous years doing R&D within the cozy reality distortion field of Apple Computer. I am currently based in Scandinavia (Reykjavik, Iceland) where I am witnessing first hand the mobile phone explosion. "Some fun facts from the mobile phone scene in Europe: * Every month more the one billion (1,000,000,000) SMS messages are sent. SMS is the neccessary precursor to WAP that educates the public to the idea of using a mobile device for data in addition to voice. It is also easy to architect Internet services to use either/or/both WAP and SMS. * Essentially every young male in his twenties in Scandinavia has a mobile phone. If you are young and breathing in this part of the world you (or one of your crew ) has a mobile phone permanently fused to one of your hands. "It is quite a treat to snoop on the distinctly non- American mobile phone phenomenon. It is spooky to go back to the Valley and see how utterly broken the mobile phone scene is (and will remain so for *years*). Weird. "As an expatriate and former Silicon Valley American, I have my own ideas on how things will play out and would like to hear what other SMART-ies have to say on the subject. There are some serious dumb/smart network and net/bell head issues that need to be resolved here. [Yeah, wait 'til the automobile-dependent U.S. public starts surfing the web and retrieving its email from behind the wheel. It'll make the tobacco companies look like good guys. -- David I] ------- From Alan Schwartz [] on The Myth of Five Nines -- SMART Letter #41: "My little network reliability story discloses a management problem. This week my phone service went out on Monday. This is my primary home number and my DSL line. Nuttin. I call Bell Atlantic. They say they can have somebody there on Wednesday. I say I bet a crew in the area messed up my line, as it is not likely that in the middle of a sunny Monday my service should go out. They say there is no way to tell where crews are, as they are dispatched by phone number not address. I express incredulity, as I assume Bell Atlantic has access to reverse directories -- after all they ARE the phone company. I hang up. "I walk around back and down the alley behind my house. I find a crew with a raised cherry picker working on the cable. I tell them my problem. Within 15 minutes my service is restored. I feel lucky I found them; and I am peeved at Bell Atlantic. 1. Without my intervention I would have been out of luck for 2 days. 2. I saved Bell Atlantic from sending out another crew. Why is this so hard?" --- From Adam Rosenberg [] on The Myth of Five Nines -- SMART Letter #41: "The telephone reliability issue is, I believe, much, MUCH worse than just the problems you discussed. "When the UNIX/C silliness spread to the telephone world in 1986, Holmdel [NJ] switched from an EPLX-based 1A-ESS to a UNIX/C-based 5E. What started then was a rash of mis-dials, incorrect dials, and memory leaks which has persisted to this day. . . . Today I take for granted that I will have to dial a call a second time once in a while. "In Texas we have "metro" numbers. Somebody's phone number will be "metro 972-123-4567" which means I have to dial it without the leading one from some area codes and with the leading one elsewhere. I don't always know which numbers are metro numbers. That means I must re- try many 972 or 214 or 817 numbers. Ack! ['Easy to use as a telephone' my ass! -- David I] "At least when the line is dead, it's dead. But the creeping meatballism of misconnects, no-connects, and software-generated wrong numbers would be more terrible if, as you say, one is having a heart attack." --- From Arthur Einstein [] on The Myth of Five Nines -- SMART Letter #41: "God, I'm glad you are good enough to 'be fair' to Bell Atlantic. We've had way more service calls than you have. The lines are so noisy my 56k modem usually gets 24k. Every night I pray for a monopoly to come along and get my phone service back on track." [Be very careful what you pray for, Einstein! She might interrupt Her dice game to grant your wish! -- David I] --- From Bill Morrison [] on The Myth of Five Nines - - SMART Letter #41: "MediaOne took 3 truck rolls to install my cable modem + phone service and three more to fix it. Local number portability (LNP) issues are alive & well." --- From Peter Varhol [] on The Myth of Five Nines - - SMART Letter #41: "Your quote: 'If I buy three services on three different physical layers from three different service providers, and if each service is two 'nines', then I should have a six 'nines' probability that one service will always be up.' "Statistically, your conclusion doesn't necessarily follow from your premise -- not without further information. [Peter raises a good point -- I am assuming that the three different ISPs are statistically independent. That is, they share no common mechanisms. Common mechanisms can have the same weak point(s), so they're prone to common failure modes. When "network effects" take hold, and Microsoft or Cisco or a given transmission medium gains dominance that approaches 'monopolistic', then systems become liable to disastrous common failure modes. In such cases, failures are not independent, so the probabilities of failure are not multiplicative. I am making a straw-man assumption that one can buy three modes of access, e.g., dial-up, DSL and mobile wireless, and access three different ISPs who use different equipment with different IP stacks and connect to three different backbones. In this ideal case, the failure of any given system is a truly independent event. This, in a nutshell, is why real, multi-party competition is critical to network reliability! -- David I] --- From Scott Baker [] on The Myth of Five Nines -- SMART Letter #41: "You're right about DSL problems. Now that I finally have my [Bell Atlantic] DSL service up and running, there are such incredibly long delays between the fast responses that it is no better than a regular 56Kb modem! It goes to sleep between bursts of activity. Bursts are very fast (over 449kbps according to one test site) but in between there are no bytes sending or receiving at all. I've taken to running multiple sessions in my browser so I can do something in one while waiting for a page to come up in another. For this I pay $49/month?!" --- Peter Cochrane [], the Chief Technologist of British Telecom, and a SMART Person, gets the last word on The Myth of Five Nines -- SMART Letter #41: "I liked this one! You hit a few hot buttons for sure - and wireless is universally crap! One point of clarification: The 99.999% is averaged over all time for all customers. I wrote a piece a few weeks ago on the dangers of averages...and you picked up on exactly some of the points I made...and worse...failures ain't random as people like to assume ...they are chaotic clustered, correlated, and causal... "In my own case I have had this phone line for 16 years and never had an outage of any kind! I hit 99.999 with ease! However - I agree - it depends where you live. In a subsequent email, Peter wrote: "You prompted me to look this up! The 99.999 refers to the up time of the switches and major network fabric and does not usually include the local loop! The USA has a significantly poorer performance in the local loop because the majority of the plant is overhead, whilst in Europe it is mostly underground. Then in his next email, Peter wrote: "It is a general design objective that the network should have an up time of 99.999 - I'll have to look up the reference . . . " In his next email, Peter enlists the help of BT employee Ian Dufour, who writes: "Tricky. The net is no help but a trawl through a few books helped. "Valdar's 1990 SPC (Stored Program Control) book (Ch 18) says, 'The stated requirement for availability of electromechanical systems was that they should not be out of service for more than 2 - 4 hours in 20 - 40 years (the exact figure depending on administration and the system)'. "As it happens 2 hours in 20 years is 6 mins a year. And Five Nines is 315 seconds or 5.25 mins a year. So the origin is almost certainly in the mists of Strowger time! It doesn't look like it was CCITT/ITU but it does look like it was switch based. "(We know local loops never achieved that level - they were typically 1 fault every 10 years but the outage time was much longer.)" ------- CONFERENCES ON MY CALENDAR September 13-15, 2000. Lake Tahoe CA. TELECOSM. Featuring George Gilder, Clayton Christensen, yours truly, and a cast of geniuses, troublemakers, and people who got rich by listening to George. This thing sells out, folks -- a word to the SMART. November 5-9, 2000. Rose Hall, Jamaica. Porter Stansberry's Pirate Investor's Ball, featuring Eric Raymond, Tom Petzinger, Porter's impressive research director David Lashmet, and yours truly. Porter is a big-picture guy, a cross between George Gilder and Tony Robbins, with a nose for leading edge values in infotech and biotech. Contact Andrea Shaw,, 410-223-2648. ------- COPYRIGHT NOTICE: Redistribution of this document, or any part of it, is permitted for non-commercial purposes, provided that the two lines below are reproduced with it: Copyright 2000 by David S. Isenberg -- -- 1-888-isen-com ------- ADMINISTRIVIA [to subscribe to the SMART Letter, please send a brief, PERSONAL statement to (put "SMART" in the Subject field) saying who you are, what you do, maybe who you work for, maybe how you see your work connecting to mine, and why you are interested in joining the SMART List.] [to unsubscribe to the SMART List, send a brief unsubscribe message to] [for past SMART Letters, see] [Policy on reader contributions: Write to me. I won't quote you without your explicitly stated permission. If you're writing to me for inclusion in the SMART Letter, *please* say so. I'll probably edit your writing for brevity and clarity. If you ask for anonymity, you'll get it. ] ** David S. Isenberg, inc. 888-isen-com 908-654-0772 ** -- The brains behind the Stupid Network -- **