I started studying the Y2K issue a year and a half ago in the interest of finding the truth about its potential impact. I took it seriously after seeing how much money the power company I work for was spending in remediating some of our equipment. I work in a nuclear plant and many people would be surprised at the scarcity of computer control, relatively speaking. We did have a couple of subsystems that were proven to be non-compliant in testing and would have brought the plant down had they not been remediated... that got my attention!

I followed the Y2K community and listened to all sides, from Gary North to John Koskinen. Still, I don't recall hearing even the worst doomers calling for anything worse than 8 to 10% of the Y2K related failures to occur on New Years Day. Much to my surprise, we made it through that day with something far less than an *% failure rate. But then I thought about it, wouldn't most of the catastrophic failures be the easiest to detect, and wouldn't they also get the lion's share of the repair/work-around efforts?

It is reasonable to think that the cummulative effects of undetected or "back-burner" problems would have the potential to really gum up the works. The longer they run undetected, the more havoc they could wreak, it would seem. I'm already seeing the effects of something unusual taking place in our society. The technological miscues I've personally witnessed this week are way out of scale!

For example, I've seen a friend have the police attempt to return his stolen car to him, the problem was, it wasn't his car -- the name and addess were his, but the car and tag info were not. What was most suspicious was the expiration date on the printout of the regristation, Feb - 1900! I've seen a non-compliant coffee pot that had to be unplugged and reprogrammed, a television that cut off at midnight, Dec 31st, and had to be reprogrammed. Ny daughter had to wait 3 hours to take her learners permit test because the computers were "having problems". My credit card company duplicated a purchase on the 3rd to my account (never has happened before). My ISP is having trouble recognizing my user name in their database and has had to make adjustments on their end twice this week (and I'm not the only one to whom this has happened with this ISP).

Are all of these things coincidental? *Could* be but honestly, I think the odds are pretty high against it. I didn't have to have my Y2K radar out to detect these things either, I would have thought them suspicious before I ever heard of Y2K.

So I now am left to wonder why it is that so many people on both sides of this issue seem to have passed final judgement on it? From most predictions I read, we've still got a good 75% of the going ahead. Is it because the initial assessment was so off base that it stands to reason the the ancillary effects will follow suite? I'm still not certain how this will play out in the financial markets; I really see a potential for trouble ahead.

-- John Cauthen (johnr@cetlink.net), January 08, 2000

Answers

John,

Good post, and your questions are well founded. I think the answers lay in the "Y2K" issue itself. It never seemed to be an issue of fact but an emotional one. No one, but no one really knew what would happen. The doomers followed Gary North's lead of "Its already too late", "no bank is compliant" and "the division of labor will collapse" - all is doom. Dr. North himself seems to have abandoned ship. Last week he posted a note on his site indicating that he was considering going back to his "one million dollar a year cash cow" business idea. Even considering that idea on his part indicates to me that he no longer feels that the banks will fail or the division of labor will collapse. This flys in the face of his (more than one) Email to me stating that I "would assume room temperature" in the not too distant future with all the other DGI's. (Not that I was a DGI. However anyone who didn't agree with Gary 100% was considered by him to be a fool and an idiot - I'm guilty, I suppose.)

So is Y2K still going to be a problem? Absolutely. Is it logically too early to tell how bad it will be? You bet. Has logic ever lead the way on this issue? Not really. Was I caught off guard with how little the initial effect seemed to be? I was floored - Italy? Go figure.

-- Darby (DarbyII@AOL.com), January 08, 2000.

Good post. What part of the Country do you live in?

-- james hyde (hydesci@gte.net), January 08, 2000.

John:

With all due respect, Y2k is over. All the problems that you have mentioned are minor issues. Coffee pots, television reprogramming, duplicate credit card bills, etc... none of this will bring down civilization. They are all minor problems that can be fixed painlessly and inexpensively. Why bother with trivial errors?

These sorts of things happened before Y2K and they will happen after Y2K but they: 1) Won't cause a recession, depression, or financial collapse 2) Won't cause a fortune to fix 3) Won't threaten lives

If we continue to have these minor problems, who cares? There's nothing out there that will make the most dire predictions come true.

Frank

-- Frank (frank@itsover.com), January 08, 2000.

Too many small leaks can sink a ship. Too many minor errors in judgement can cause an airliner to crash and burn. Too many small leaks in the dam can cause it to burst. Do not dismiss the small glitches too lightly. So many of them, added to the natural glitches, and then have a magor one or two, and buddy you have a real problem even the spin masters will not be able to hide.

-- Notforlong (Fsur439@aol.com), January 08, 2000.

Here's your answer.

Suppose, as you've suggested, that we've seen 25% of the problems so far. You correctly point out that that leaves 75% of the problems yet to come--a huge magority, in fact.

Great. Take every problem reported to date, and multiply it by three.

Pretty scary, huh?

And, just to get a really accurate picture, spread _that_ number of failures out over the course of the next twelve months.

Bottom line is: that level of "disruption" will scarcely break the static threshold of failures and problems that happen every day. Three times nothin' is still nothin'.

Maybe if we've only seen...oh...say one problem out of every five hundred, or one out of every thousand, we'd have something to worry about. But that would mean that fiscal year rollover, lookaheads on 1/1/99, lookaheads on 12/1/99, the rollover itself, and the first workweek after rollover, all together, had only exposed 0.2% or 0.1% of all Y2K failures. How likely do you think that is?

It's over. The odds of any substantial Y2K disruption --one that would require you to fire up the generator and open that tuna fish-- are essentially nil.

Craig

-- Craig Kenneth Bryant (ckbryant@mindspring.com), January 08, 2000.

Let it go people,

Minor glitches are occuring, thats it. Take your preparations, store them for a rainy day.

Myler bags, oxygen absorbers and 5 gallon buckets will ensure most of your food will last for a long time.

Put a cover on your generator, do routine maintenance on it. If you have alternative engery? Keep up the maintenance.

Since the rollover, I've slept like a baby. I've finished all my projects in preparation. It didn't happen. My family is safe.

I can sleep at night now knowing that the lighbulb that burned out in Bucksnort, Tennessee because of a Y2k glitch won't have any affect on my families health, care and welfare.

p.s. Keep good records while using your credit card. Y2k must of prepared you to ensure all your finance records are updated. Why worry any longer? If something major happens? Your ready. cork

-- cork (corcorab@hotmail.com), January 08, 2000.

John

You seem to be a y2k magnet!!!!

I don't know how old you are...old enough to remember the firts computers??? There were days when bills were printed on punch cards to that they could run it through the machine again if needed. They used to have instructions not to "punch, mutilate, or fold"...in those days, people were suspicious of computers...

Now, we all love them...if y2k shows its face in a dozen small ways, I can see a lot of people losing faith in computers...

Can you imagine a world where almost everyone (lawyers included) can say..."can you PROVE beyond a shadow of a doubt the integrity of your database""

Great adventure ahead...little to threaten life but many to tell about over coffee.

-- thom (thomgill@eznet.net), January 08, 2000.

Guess I am too paranoid about that coal filled train cars, sitting idle on the track, without an engine. I do not remember seeing this before, not that I was ever inclined to notice. My electric company, depends on coal, My first wonder at those silent box cars, could not, someone come in and steal the coal? Where are the coal police?

-- Saw Boxcars (Heavenhelp@usall.com), January 08, 2000.

The truly wise are 100% Polly.

And 100% Doomer.

But most of us are incapable of this seemingly impossible paradox. And so we must oscillate with the shifting tide of public events and private moods. Overreaction, underreaction. So the wheel turns, fed largely by its own imbalance.

Did the Doomers overreact? Only if they *failed* to prepare for the most difficult future of all: Normalcy.

Do the Pollies under-react? Only if they fail to accept that governments, societies, and the individuals which make them up who intend to survive over the long term, must be on guard for the worst of crises.

Crisis and normalcy. Aren't these *together* the two sides of life? Has any of us experienced a life wholly without both?

To the Doomers: It's never too late.

To the Pollies: It's never too early.

To the Prudent: Everything is nearer than you think. Learn how to open your eyes. Learn how to stay relaxed. And learn how to put acquiring real friends before the acquisition of anything else.

For every thing, there is a season. In real life, timing is everything. And everything has its time.

-- tim phronesia (phronesia@webtv.net), January 08, 2000.

For all we know, these estimates remain pretty accurate. We might indeed have suffered 8% of failures at rollover. And somewhere around 25-30% of all failures were supposed to have happened before rollover according to Gartner, although Hoffmeister has made a very solid case that Gartner's numbers did NOT include second-order problems (that is, problems associated with efforts to AVOID date bugs, such as new system implementations). Hoffmeister argued that problems with such switchovers would be the worst problems of all (such as Hershey, Whirlpool, etc.) and so far this has been correct.

By today, we've encountered about half of all actual problems according to Gartner, and maybe 75% according to Hoffmeister. And the rest are supposed to be spread out over more than a year, therefore occurring at a rate which will not exceed our ability to adapt.

This leaves plenty of opportunity for individual snafus, some of which might be even more devastating to the economy than Hershey (the worst yet). Unfortunate for those in the immediate line of fire, invisible otherwise.

-- Flint (flintc@mindspring.com), January 08, 2000.

THE y2k issue was always, will continue to be, data. When the data steps outside the parameters allowed by the programmer, the behaviour of the program(chip, plc board, et cetera) will go into 'undefined' state. Some of these behaviours will manifest in the real world as problems. Some will lie hidden until some data exchange or other activity (backup, restore, year-end analysis) causes the program to exercise the corrupted data.

So the problem with y2k is more sophisticated than merely a date. or even a group of dates.

Remember: it is when the corrupted data is called upon that the problem *could* appear. Such things are sort of predicatble for various classes of programs. Thus, you might not ever see a problem if you never have to restore from backup. Or you might see a problem the first time that you run payroll checks. Or the first end-of-month processing. Or the first batch reconciliation of cross company charges. So...for the business (enterprise class of programs a 'la cory) systems, we face predictable points at which undefined behavior might manifest. At each of these points (payroll, end of month, end of quarter, et cetera) the enterprise systems will HAVE to be revalidated. You must continue to check in order to make sure that you do not have a creeping corruption to your enterprise data sets. And you must check after each and every processing period.

If you do not, then how will you know that you have JUST caught the data corruption?

Bear in mind, especially for the internet companies, the data set of the firm is frequently its true asset.

Other factors contributing to data set integrity can also be calculated. The total number of inputs to the enterprise software constitutes its universe of vulnerability. These input points also have their own points of vulnerabililty and processing cycles. These have to be checked as well.

Otherwise you just don't know.

And you need to have designed a test program to validate the data. Otherwise you just don't know what you are testing.

-- pliney the younger (pliney@old.bald.guy), January 08, 2000.

pliney:

Yes, these things are all possible. But stating valid possibilities gives us no clue as to their likelihood. I think most pessimists are surprised by the lack of difficulties because they confused what is possible with what actually happens in practice, and concluded that every bad thing that unquestionably *could* happen, therefore inevitably *would* happen.

Somehow, I doubt that remediators overlooked processes that have yet to be performed. Testing usually involved both simulated and real data. And except for infrequently run processes (like end of month or quarter), stored data now being recalled and used probably includes damn near all of it. Not many places left for subtle data corruption to hide anymore.

-- Flint (flintc@mindspring.com), January 08, 2000.

tim phronesia,

Really enjoyed your post. Y2K (and the polarizations that have accompanied its unfolding) is as much a philosophical issue as a technology event, it seems.

-- (RUOK@yesiam.com), January 08, 2000.

Could you give us the makes and model numbers of the malfunctioning equipment? (The coffee pot and TV)

Mikey2k

-- Mikey2k (mikey2k@he.wont.eat.it), January 08, 2000.

Wow! I didn't expect my question to elicit so many responses, so quickly. Not only that, very well reasoned and thought-out responses! There is hope for civil discourse after all!!

Let me respond: *So is Y2K still going to be a problem? Absolutely. Is it logically too early to tell how bad it will be? You bet. Has logic ever lead the way on this issue? Not really. Was I caught off guard with how little the initial effect seemed to be? I was floored - Italy? Go figure.

Darby, I agree with your comments about Gary North, he does seem to have an "end of the world" agenda, or at least he did. On the other hand, he came off well on the radio when I heard him on Art Bell. I can see how he got so many people in his camp. I credit him with putting me onto some good links, but I also saw a fatalist spin in many of his editorial comments about them. To his credit, he at least separated the editorial content from the source content.

I too was a bit shocked over the lack of severity of the initial impact. I expected many water and power problems in Southeast Asia, Russia, and Italy. The emotional tone this issue took on was what really shocked me. We had no precedent for dealing with a problem of this sort, yet we had experts claiming they had the absolute truth about its impact on both extremes.

It got to the point where "Y2K" precipitated the same reaction as three other alpha-numeric characters - "UFO". Mention either of them in a serious breath in most circles, and you're looked upon as though you'd flatulated in church! Again, another puzzling phenomena.

*Good post. What part of the Country do you live in?

James, I live in the southeast.

*With all due respect, Y2k is over. All the problems that you have mentioned are minor issues. Coffee pots, television reprogramming, duplicate credit card bills, etc... none of this will bring down civilization. They are all minor problems that can be fixed painlessly and inexpensively. Why bother with trivial errors?

Frank, Y2K may well be over for all I know. I'm not a computer programmer, so I really don't have a technical feel for what's happening. I have to wonder when these problems will have run their course. They may seem like minor problems now, but can they snowball and get worse, and perhaps more frequent? At some point, they may reach the point where they are no longer trivial. Also, are all repairs going to be inexpensive and easy?

*These sorts of things happened before Y2K and they will happen after Y2K but they: 1) Won't cause a recession, depression, or financial collapse 2) Won't cause a fortune to fix 3) Won't threaten lives

I wish I could be this absolute in my appraisal of this situation, but I just don't have enough command of the facts to do it - either way. I could see financial and administrative problems reaching a level where a recession could result. We live on such small margins now that it wouldn't take much tip the balance. How can we be so sure of the costs? The number of solutions and their complexity would seem to be endless.

Won't threaten lives? I'm not entirely certain that it hasn't already cost lives. I won't to see the final report on the Norway train crash and the Indian transformer flash. I know this sounds a bit cynical, but I don't rule anyhting out until I have sufficient data to do so... I'm suspicious of both of these incidents.

*If we continue to have these minor problems, who cares? There's nothing out there that will make the most dire predictions come true.

Frank, these "minor problems" could add up - I know I don't want them. You may see them as residue, but they may be precursors for all I know. I never thought we'd see the most dire predictions come true, however, I think it might be a bit premature to say it's under wraps.

Still, it's reassuring to hear someone confidently state that it's over - and mean it!

Notforlong, your analogy about small leaks sinking a ship is fitting. It's amazing to see the number of leaks that are occuring right now, but most have been conditioned to dismiss them as "normal failures". I can't prove whether they're normal or not, but I can say that my own little world is in more technological chaos than it's ever been, I just hope it's going to get better from here.

Craig, you mention multiplying the present problems by 3. I'm not sure that would even get the attention of the mainstream. I think it's going to take more than that. If the "Pollys" are correct, we won't be seeing any more significant bug reports come March (when the quarterly reports are compiled) than we are now.

Craig adds:

*Maybe if we've only seen...oh...say one problem out of every five hundred, or one out of every thousand, we'd have something to worry about. But that would mean that fiscal year rollover, lookaheads on 1/1/99, lookaheads on 12/1/99, the rollover itself, and the first workweek after rollover, all together, had only exposed 0.2% or 0.1% of all Y2K failures. How likely do you think that is?

I don't know Craig, this requires a lot of speculation. I just have this instinct that's telling me that some computer systems are running and compounding corrupt and incorrect data, well beyond what's normal. I don't know what the final result of that could be and I can't prove it's really happening.

*It's over. The odds of any substantial Y2K disruption --one that would require you to fire up the generator and open that tuna fish-- are essentially nil.

I agree that the tuna and generator scenario is moot, but is it really "over"? I'm staying out of the stock market until I'm sure it is :-)

Cork, how many minor glitches does it take to become "major". Are we currently seeing the worst of what the Y2K bug will have to offer? I hope so.

Thom writes: *You seem to be a y2k magnet!!!!

Not hardly :-)

*I don't know how old you are...old enough to remember the firts computers??? There were days when bills were printed on punch cards to that they could run it through the machine again if needed. They used to have instructions not to "punch, mutilate, or fold"...in those days, people were suspicious of computers...

Not that old, but I do remember the Atari 800 computer!

*Now, we all love them...if y2k shows its face in a dozen small ways, I can see a lot of people losing faith in computers...

I'm trying to learn to be less reliant on them; they were already causing me enough grief thanks to Windows XX!

*Can you imagine a world where almost everyone (lawyers included) can say..."can you PROVE beyond a shadow of a doubt the integrity of your database""

I'd rather not!

*Great adventure ahead...little to threaten life but many to tell about over coffee.

I don't see anything life threating (on a wide level), but I do see *potential* economic problems ahead.

Heavenhelp writes:

*Guess I am too paranoid about that coal filled train cars, sitting idle on the track, without an engine. I do not remember seeing this before, not that I was ever inclined to notice. My electric company, depends on coal, My first wonder at those silent box cars, could not, someone come in and steal the coal? Where are the coal police?

This is worth following. I can't explain it myself, but I am picking up signals about problems with rail traffic. My father is a railroad retiree and he's confused by what he's seeing.

Tim, your paradox almost makes sense. I'm not at either extreme so I can't fully identify with the rationale. Where I don't see your hypothesis holding up with the never too late/never too early scenario is that I see many doomers throwing in their towels after only a few days. The pollys are convinced it was over Jan. 1.

Flint, Your argument seems plausible. Gartner and Hoffmeister are credible sources. If the problems are indeed spread out over a year, they may very well be manageable.

Pliney, you really seem to have a handle on this. Let me ask you a question about one of the "glitches" I've observed. Could the problem with the registration printout the police presented to mey friend have been a genuine Y2K bug? IOW, could the 1900 year in the expiration date have caused the database to mismatch the owner and vehicle?

Thanks for the explanation of how computers process data, it makes a lot of sense!

-- John Cauthen (johnr@cetlink.net), January 08, 2000.