The problem is large and widespread. The solutions are costly and difficult. Even the realistically optimistic estimates predict that 10% of key systems will fail (which would trigger an economic cascade effect that would in itself have dire consequences). Other people paint a direr picture. Ed Yardeni, respected computer guru and author of Time Bomb 2000, expects 50% or more of computers to crash if immediate and concerted action was not taken -a prediction he made in July 1998. "Infomagic", the anonymous programmer whose reports I have included here, has mathematically predicted what could be in effect a return to the Dark Ages. Infomagic's views, it must be added hastily, are seen as pessimistic by most- but it's hard to argue with the raw facts of his mathematics.

The problem here is that we live in a very complex, very interdependent society. If one component ceases functioning, then the things that depend on it will also stop functioning. And thus the things that depend on each of those things will stop functioning, and so on. As this happens, people -being people- will panic, which will in turn make things a lot harder. Also, some factors -such as power- are so essential to our society that if they alone went down for a significant period of time, serious chaos would erupt. A week-long blackout across the entire United States would not be an enjoyable proposition, given the record of "events" (massive increases in crime such as looting, robbery arson, even rape and violence) from previous urban blackouts. Couple this with a situation where the emergency services are unavailable (as they would be if the phone systems didn't work) or simply overstressed, and add to that factors such as errors in government computers (not sending out welfare cheques, for instance; far lesser things than this have caused major riots), and we have a situation in which to say urban life might be dangerous, would be an understatement.

The above does not even bring in the fact that the world economy runs on interlinked computers, most of which are uncompliant and a good proportion of which will not be compliant when the time rolls around. If these computers fail, billions of dollars will disintegrate into thin air. Asia and Europe (Asia is too concerned with their immediate economic problems to have time for this future problem; the computer people in European banks will be busy until YYY1 with the Euro) will be the most vulnerable. If Europe crashed at the same time as Asia's recovery abruptly halted, and given the current turmoil in Malaysia, Indonesia and much of Indochina, the result could very well be a world economic disaster even if every single banking computer in the United States was fully 100% compliant.

Remediation Needless to say, the various agencies of the United States government do not want this to happen. All of them have embarked on various remediation attempts, under the general control of "year 2000 czar" John Koskinen. However, the only agency expected to complete remediation by the deadline of 1/1/00 is the YYY3. The expected dates for the others range from June 2000 to 2032!

Corporations are also concerned about y2k, and are treating it as a serious problem. Boeing expects to spend $800,000,000 on it in 1999, for instance. Most of the major banks are spending equivalent sums (Chase Manhattan Corp is spending $363,000,000), and so are the telecommunications companies (AT&T is spending $900,000,000 -triple the estimate it gave earlier in 1998.) These costs, as AT&T has found out relatively early, can often be minimal estimates as more lines of code are found. One American welfare agency thought it was almost compliant..before finding another thirty million lines of code that had to be checked.

The problem is compounded when you realise that all of these compliant systems must also be tested. On average, even the most competent debugger will only flush out 90% of the bugs on his first "run". The times given to be "compliant" generally do not involve the time taken to test for more bugs, and to fix them. The testing, further debugging, then more testing and hopefully the final debugging, could double the remediation times. Testing itself will be difficult in the cases of the big corporations and essential-services-agencies (such as power or water), who will need to keep their primary computers running 24/7 and thus will have to rent powerful mainframes to carry out their testing on. With thousands of major companies carrying out such testing (and initial testing can take as long as four weeks), there will be a shortage of such mainframes, leaving many companies unable to carry out alpha testing- let alone the alpha remediation, beta testing, and beta remediation needed for a program to be verified mostly bug-free.

And to make remediation yet more difficult, the year 2000 bug exists, as has been discussed, in older languages, primarily and specifically COBOL (COmmon Business Oriented Language). COBOL is not something that the generation of programmers coming through colleges now learn a lot of. Neither is it something that can be learned easily or quickly; at least not to the extent of being able to effectively debug someone else's program without inadvertently writing more errors than you remove. In short, there is a finite number of skilled programmers, a finite pool that is getting smaller as many programmers realise the severity and the danger of the problems that y2k could cause, realise the odds against fixing them, and "bug out". Demand for skilled COBOL programmers is also going to go sky-high in 1999 (in the last few months, the hourly pay rate for experienced programmers has already doubled) and it is possible that companies have not anticipated this in their budgets. Programmer and y2k expert Ed Yourdon predicts that by September 1999, such programmers will be asking for -and getting-$600+ per hour (about ten times what they currently recieve).

Remediation of the 25 billion embedded chips is, of course, impossible. They must be replaced. Some embedded chips (in satellites or oil wells, for example) can not be replaced. Many others will be extremely difficult and costly to replace (for instance, taking apart an engine to remove a specific chip in an internal component, replacing it and putting the engine back together...five million times). Replacing these embedded chips will be a time-consuming, extremely-- cost-consuming (in some fields, such as kitchen electronics) procedure. Only one automobile company, as of December 31 1998, has done even a limited recall because of embedded chips. Although only 3-5% of embedded chips are expected to go wrong, three percent of twenty-five billion is a massive amount -and few people know exactly which chips will go wrong.

The earthquake is coming. We know the exact date it will hit: 1/1/00. What we do not know are the consequences. Optimists predict a cost of around $1trillion and perhaps only a few dozen deaths due to it. Lesser optimists give convincing evidence that the cost could be in the order of tens of trillions, and the deaths of thousands or tens of thousands. Pessimists talk about "The End Of The World As We Know It", incalculable damage to or destruction of property, and hundreds of millions of deaths.

What will in truth happen, nobody knows until it happens. There is little positive evidence. There is a lot of negative evidence. The only definite decision is that it would make sense to prepare.

Opinions please?

-- Leo (leo_champion@hotmail.com), December 16, 1998

Answers

graph 1- - "direr" maybe "darker"

On balance (which this IS) not a bad shot. You need a good editor to tighten it up a bit but, all in all, not bad. You could probaoly go with it as is, and get the reader VERY interested in whatever you wanted to sell as long s it's something to help him/her over the rough spots coming. Since I understand what you are selling, it'll work!

Chuck

ps Now, all you need is the location, the construction crews, and the product buying team.

-- Chuck a night driver (rienzoo@en.com), December 16, 1998.

Leo,

I'm sure you know this already, and you were just tired when you typed that up, but it's *Ed Yourdon* who's the computer guru and author of "Time Bomb 2000".

And, of course, *Ed Yardeni* is the well-respected U.S. economist who correctly predicted the bull market of the mid- to late 1990's, who now says there is a 70% chance of a nasty recession in 2000 because of a Y2K caused information shortage. (Similar to the 1974-75 recession caused by the oil/energy crisis of the time).

-- Kevin (mixesmusic@worldnet.att.net), December 16, 1998.

Leo,

Here's a newsgroup posting with very up-to-date links to information on just how deadly serious Y2K is. If you're looking to write a TEOTWAWKI scenario to convice clients, this one link will give you all the arguments and links you need to make your case:

http://x2.dejanews.com/getdoc.xp?AN=421054011&CONTEXT=913682332.301858 889&hitnum=1

"y2k retort: my letter to a pollyanna"

Leo, I expect a check in the mail by morning. That, or promise me you'll go out and buy some toilet paper, water, and blue tip matches FOR YOURSELF this weekend... 8-)

-- Kevin (mixesmusic@worldnet.att.net), December 16, 1998.

Leo - first paragraph third sentence - wrong Ed.

Knock it off about COBOL please. School I went to required COBOL at least through 87 or 88, still was offered through 94 or so. That is not that far back. And COBOL is very easy to learn and use if you know any structured programming language. So the shortage is more of a myth than a reality - there is a severe shortage of COBOL programmers willing to work for $25,000 per year and no benefits - which I was offered a couple years back. Needless to add I laughed at the guy - guess he is still looking.

-- Paul Davis (davisp1953@yahoo.com), December 16, 1998.

Those two changes made: Yourdon replced Yardeni, and I killed the bit about COBOL not being taught in colleges.

Chuck, will fix that paragraph now. My worry is that I don't want to sound like an extremist predicting the end of the world. I'm being very optimistic here..and it still sounds like doom.

Still, it probably can be done. And probably should.

Thanks for the advice.

--Leo

-- Leo (leo_champion@hotmail.com), December 16, 1998.

Change recieve to receive

Leo, it's time to rig your own boat. What will you do with the money you hope to make, when what you need is no longer on the shelves?

I get the impression you're not really convinced, yourself, even though you "understand" the situation. I suspect most here are personally acquainted with this phase. Certainly I am.

-- Tom Carey (tomcarey@mindspring.com), December 16, 1998.

Ed Yourdon--not Ed Yardeni--wrote Time Bomb 2000. As a long-time newspaper editor and publisher, your piece does need some editing and tightening.

-- Vic Parker (rdrunner@internetwork.net), December 16, 1998.

Leo,

Note corrections already mentioned above.

-- You may want to consider having one or two pieces that are sent to your target market prior to this one. Although I think you have a good handle on the situation, there are many prospects that are going to dismiss you after reading this, unless you bring them along a little bit slower. I've talked with lots of small business owners, and the bulk of them will assume you're just hyping them for the sale.

If I cold call a prospect (either by phone, mail, or personally) and start the conversation on Y2K with anything close to what you have above, they will put up a brick wall. At this point, they assume you are feeding them a load of BS in order to sell them. Or worse, they assume you're a crackpot. With most people, you generally have to "ease" them into your views. For instance, I first ask the person if they've considered Y2K and its impact on their business. Usually, they answer no, or they think it won't affect them. I'll comment that it sure looks like it could be a big pain. They ask why. I respond with, "Well, everybody depends on computers so much now." They'll usually agree to this, and then comment that their accounting dept. etc. couldn't get by without computers now. Then I might make an off-the-cuff comment about "Wonder what would happen if it does turn out bad? You know, people can't get Soc.Sec. checks, Medicare, etc.?" The person normally responds that you'd see a lot of angry people, etc. It's usually at this point that the person you're talking to will begin to question you about what "could" happen. Spend the next 2 hours talking with them, and they're convinced.

Note: I'm not advocating deceiving people; just that if you want to persuade someone, you have to guide the conversation rather than control it. Most people (including myself) tend to think their own ideas are better than anyone else's. Not that any of us are correct in our assumptions, it's just the way it is. Lead them, don't push.

-- Greg Sugg (gregsugg@bbnp.com), December 16, 1998.

Tom, I LOVE a man who can spell!! :-)

Leo, it's not that big of a deal, but the word is actually incompliant, not uncompliant. (PS- I love your enthusiasm!)

-- Gayla Dunbar (privacy@please.com), December 16, 1998.

Leo, good idea to write a wrap-up article, but it will be easier to read if you give it an outline structure using headings, table of contents, etc. Then publish it in HTML to receive widest possible distribution. For a handout pamphlet, condense it and print an URL to the full version.

Also, drop the 'earthquake' metaphor, it does not properly convey the pervasive and insidious nature of the y2k bugs. A better analogy would be to a mutating virus such as HIV which can propagate through system interconnections and can remain occult for long periods of time.

And yes, we are talking about bugs, many varieties of widespread programming errors, the problems compounded by varieties of languages and varieties of date formats. Properly speaking, we are looking at more than just y2k bugs per se, we are facing a calendar problem which includes errors relating to the millenium digit rollover, but the calendar calculation problems will manifest throughout next year (starting in 2 weeks!) and will continue long beyond the year 2000.

So don't talk about 'the exact date it will hit' it is not the kind of problem that hits on a specific date. This may lull people into thinking it's remote. Mention that there are only about 50x5=250 business days until the rollover. Mention that not only is it difficult and expensive to remediate software systems, but that it is impossible to complete. Also mention that not only is full remediation of embedded systems impossible, but replacement of all these systems is impossible to complete as well.

Establish the fact in no uncertain terms that there will be system failures, and the issues for consideration are the extent and severity of the failures, along with what should the average person do about this both now and throughout the next three years.

The incidence of system failures is more likely a gamma distribution, which is similar to a normal distribution --the bell curve-- but with a slower taper after the peak, and which may be centered on mid-2000. The failure curve may be expected to rise further next month. The SF power outage was just an outlier data point on the leading edge of the lower tail. Gamma distributions are commonly used in modeling failure rates and survival risk.

And in mentioning death rate increases, the reader will want to know what causes this (starving, freezing, fires, home invasion robberies, ...?).

-- Jon (jonmiles@pacbell.net), December 16, 1998.