"Endgame" essay preaches only to choir

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

Yourdon's essay, if nothing else, indicates that he's becoming as opinionated and shrill as I am. And as a result, he seems to be wandering off into the deep end without realizing it.

The whole essay uses a chess game as a metaphor. This is very instructive, because what's critical is that a chess game has a defined, clear winner and loser. And this isn't a trivial observation either, because Yourdon relies throughout the essay on winning and losing at the y2k game, as though they were as clearly defined as in chess. Yourdon is more subtle than most authors, but in the end, the implication is that your bank, your town, your utility either "wins" because its y2k work is utterly completed, or it "loses" by becoming entirely nonfunctional. No bank, no power, no police or firemen.

Once this metaphor is established, Yourdon uses it to expound on his favorite topic -- how software projects don't come in on time, how nobody meets his satisfaction with their current status, and how many significant economic actors are hopelessly behind and will "lose" the endgame. If you buy into the all-or-nothing nature of the chess metaphor, it all seems to add up.

But buried in the middle of this long essay is an interesting sentence. Yourdon is talking about the track records of the actors, and rhetorically asks how much confidence you'd have in a company that had screwed up every software project they'd touched for 20 years. The "right" answer, of course, is that this company will "lose".

What's interesting is, companies like this actually exist! And Yourdon knows it. Just for fun, let's turn this same question around a bit. If a company has screwed up every software project they've touched for 20 years and are still successful, and y2k is just one more software project to screw up, what is the likelihood that this company will soldier on just fine after screwing up one more? According to the metrics Yourdon is so fond of citing, most big software projects "fail", yet the companies suffering this fate continue on just fine.

Maybe a better metaphor for remediation activities would be trying to negotiate a big raise from your boss. You ask for 20%, and he counters with 2%. If you end up agreeing on, say, 10%, did you "win" or "lose" this game?

I don't consider this an idle question at all. If your bank sends you a statement with the year as 1900 but the amount correct, did they "lose" the chess game? If your electric utility skips a few bills and then hits you with one big one for 3 months' worth, but the power was up all that time, did your utility "lose"? If every stoplight in town goes to blinking red for a day (with big lines everywhere) before it's corrected, did your town "lose"? If some of GM's assembly lines are down for a day, or WalMart can't restock some orders for a week longer than usual, or gasoline prices rise 50 cents a gallon for 6 months before dropping back down, are these all "losses"?

Read what Yourdon says about banking. He can find only two problems -- that bad data may spread corruption, and that there might be cross defaults from "noncompliant" customers. But in real life, a significant level of testing has effectively ruled out the "bad data" boogyman, and noncompliance does not equate with death (or inability to repay a loan) any more than a 20 year history of software incompetence has meant this in the past. So Yourdon is building his case quite clearly on policy statements both contrary to observation. But he's careful not to mention this -- apparently it undermines his black-and-white world, for purposes of this essay.

Anyway, what I'm trying to say is that Yourdon's fundamental assumption, that y2k remediation is all-or-nothing, and all the players either achieve perfection or "lose" and die, is incorrect. He's subtle about it, but still wrong. Instead of reading this essay with the TB2000 Will To Believe, try reading it with a critical eye to simplistic assumptions. And it doesn't hold water.

-- Flint (flintc@mindspring.com), September 07, 1999

Answers

Flint:

I think what Yourdon--and many on this forum--fear(s)is simultaneity.

Sure, a lot of companies have screwed up every software project for 20 years, but virtually all of them happened absent the same thing occurring on a multitude of fronts.

If you project simultaneous failures across the broad spectrum of our economy and the world's economy, the chess analogy isn't farfetched at all, because it does raise the spectre of the "end game."

As for preaching to the choir, of course he is. You and a few other notables are the only ones who aren't. We still listen, though, even if you are slightly off-key.

-- Vic (Rdrunner@internetwork.net), September 07, 1999.


Flint,

You make many good points.

However, you commented:

"What's interesting is, companies like this actually exist! And Yourdon knows it. Just for fun, let's turn this same question around a bit. If a company has screwed up every software project they've touched for 20 years and are still successful, and y2k is just one more software project to screw up, what is the likelihood that this company will soldier on just fine after screwing up one more? According to the metrics Yourdon is so fond of citing, most big software projects "fail", yet the companies suffering this fate continue on just fine. "

Not necessarily. A company may have come in late on all their projects for the last twenty years and still be fine because they could always continue to rely on their existing system until the new one was in place and functioning. Not so for Y2K. If they aren't done by the rollover, they won't be able to operate until they get their new Y2K compliant system up and running.

Take the FAA for example. Over the last year and half, they have been plagued with problems. Every time they try to install a new radar system, it malfunctions, and they have to revert back to the old system. They won't have that luxury next year. They know that the existing equipment won't work in the year "00". Recently, the FAA publicly acknowledged that they aren't ready and the project is years from completion. If this is true, what will the impact be? (Remember, the FAA said it, not me). This is not a minor point.

I suspect the same is true for many of the other organizations that are supposedly "on track". What would happen to a major corporation that found itself in a similar situation? If they knew that they wouldn't have critical systems in place for several years to come, they are out of business. They can continue to run their legacy systems for the rest of the year and lie to the public until the end (giving the company executives time to cash in on all those stock options), but the fact is, even now, the company has already "lost".

-- Clyde (clydeblalock@hotmail.com), September 07, 1999.


I find the "moderate Flint" to be a thoughtful, aleit optimistic GI. The "devils advocate Flint" keeps surfacing and confuses me a little. Yourdon's analogies have to be simple given the need to hit a wide audience. (In contrast, Cory's writeups require a degree in computer science and two six-packs of Jolt Cola to understand.)

The big difference between the failed software projects that you refer to and those upcoming failures, in my opinion, is that most software packages in the past were, to varying degrees, optional. A new billing system? Heck. If it failed, you used the old one for awhile longer. In this awkward fashion mixing failures and successes, slowly but surely the global computer system evolved alot like biological systems evolve.

Let's carry this biological metaphor a little further. Complex systems aren't born, they evolve concurrent to gradual changes in the environment, never straying too far from equilibrium. Due to the miracles of population genetics, the temperature drops a fraction of a degree, the population puts a few ounces on their bodies (i.e., Eskimoes). Once in awhile, however, something big happens. Certainly, the celestial body of 67 million BC hitting the earth was one of these. More recent examples include the introduction of a new species into a niche. Tree snakes introduced to Guam have now eaten all of the birds and rodents in short order. When such big events occur, the system is purturbed rapidly and substantially from equilibrium. At this point every organism faces an imperative: adapt very quickly or die! I think that y2k could prove to be such an event for organizations world-wide. The time required to make drastic changes necessary for organizations to adapt is very short.

PS - I'm guessing that you would find recent brief discussions of "chaos theory" and y2k interesting as well. The notion: small perturbations from equilibrium can lead to very large and unpredictable events.

-- Dave (aaa@aaa.com), September 07, 1999.


Clyde. You're fast on the draw.

-- Dave (aaa@aaa.com), September 07, 1999.

Flint,

No offense, but do you have a definitive answer for anything? You have to be the best fence-sitter I have ever seen. You are better than the teachers at the high school my children go to, at not answering a direct question.

Does anyone else see this phenomenon in their daily life? Where I work there seems to be "no wrong answers" and everyone is working on their "self-esteem". It is sort of like the kid's soccer games where they do not keep score so there will be no losers...

Like I said in the beginning of this note, no offense to you personally Flint, but I had to get that off my chest.

So, how 'bout them Cowboys???

scratchin' at the door...

The Dog

-- Dog (Desert Dog@-sand.com), September 07, 1999.



Flint, Perhaps you didn't understand the part about each company, entity, etc., is going to be evaluated by an individual. He may or maynot understand the interdependent connections between what he is viewing(say, his Electric Utility) and what he cannot see (his Electric Utilities dependency on many other entities). Either way, as time closes and the chess board looks like it could go either, any hint that your Side is going to loose could create a HUGE reaction. The public panic could be even MORE hightened because they will suddenly realize they've been wrong all this time - and now there is NO TIME.

Chess, by the way, can come to a draw or stalemate. That would be acceptable to me. Again, the simply act of preparing SOME seems so trivial compared to not doing anything. way

-- Gregg Abbott (g.abbott@starting-point.com), September 07, 1999.


Flint, you have a point here. Your ability to see shades of grey is a welcome attibute here. However, if you'll allow me to b.s. for a moment, I believe that there are situations where catastophic, systemic failure in our economy/infrastructure are possible. You say that some y2k errors will occur even though most have been corrected. Okay, I'll grant you that.

However, what we might see is that a lot of smaller y2k errors add up to a huge mess that could seriously topple things for awhile. A "Death By 10^3 cuts," as they say. This could be an avidity issue. That is, a large number of small monkey wrenches thrown into the Widget could be just as bad as a single, large monkey wrench thrown in. A large number of lower-weight events could have just as much an effect on the model as a small number of high-weight events.

This type of failure is similar to a victim of anaphalactic shock. A week ago he was fine. Now he's dead. For some strange reason, his body reacted to an infection in slightly the wrong way. Superantigens from the bug he had set his immune system on fire, sending this victim into progressively higher fever, which messed up his immune system even more, which then began to kill off the properly-functioning cells needed to mount a specific response to the bug, which in turn secreted still more pyrogens, which then began to damage the hypothalamus of the brain, which sent his fever way past 106, putting him in a coma, etc.

Cascading types of failures of this kind employ feedback. There is a point where errors loop back to make more errors, at a rate that happens too fast for the guy who mans the "off switch". Over time, the imact of failures looks like an "S" on its side, not being linear. If you really believe in BITR, you may (1) think that y2k will not stress the macro-system to the point of inflection, where suddenly everything changes for the worse. Either that, or (2) you think that the number of failures has a 1:1 linear corellation with macroeconomic impact. I think (2) is out of the question, personally.

(1) is possible IMO if y2k were ONLY computer errors, but it's not only computer errors. Y2k is a black hole of uncertainty on the economic map; no one's ever seen around the "event horizon" of the black hole to optimally plan for it. This black hole--something which has no precedent in the models used by management and investors--means a large drop in confidence because everybody's going to try to be stingy and conservative AT THE SAME TIME. The more people that think this way, the bigger the black hole gets. Pretty soon the whole economy can be sucked in. This fact will not make matters any easier for y2k repairs. This nuance is what could easily shift the y2k errors past the inflection point of no return.

This is where Ed Yourdon could be correct. At any rate, I suspect y2k is going to re-write lots of future economic textooks. Will the outcome be in favor of decentralization/free market or in centralized/ Keynesian/Eurosocialist practices? Who knows...

-- coprolith (coprolith@rocketship.com), September 07, 1999.


Flint: Yourdon's comparison to a chess game is imperfect but he seems to have been striving for simplicity given his audience is not at the genius level. However, look at Linda's extensions on the original "endgame" thread where she compares the situation to many, many chess endgames being played simultaneously, except each move is interdependent on many other moves in many other games with a wide variety of skill amongst the players. The situation is a hell of a lot more complex than you are making it out to be, and problems that complicated are not going to result in mere inconveniences as you suggest.

-- a (a@a.a), September 07, 1999.

Flint -

You are wearing thin. Time to find a new identity somewhere. The only thing you said was 'on the other hand, what if..' Yourdon is talking stats based on reality. Failed projects are simply a drain on corporate monies not a danger to their operations. These companies have enough systems and make enough new purchases to keep themselves together. Its the internally developed systems development which fails and these companies were functioning prior to the project and function after the project demise because the project was never used in the production environment which makes the company viable.

Let's go through it again.

Company A has a working environment probably purchased.

They decide to build a system in house for competetive advantage.

The business is running on the standard systems during project lifecycle.

Project fails.

Business continues to use old apps and possibly buy some add ons.

Business continues regardless of the project failure.

Get it??

-- ..- (dit@dot.dash), September 07, 1999.


Flint:

I would be interested for you to give us a single scenario (other than an electrical grid failure on a national scale) in which Y2k or any other "systemic perturbation" WILL result in a catastrophic or unrecoverable failure of our computer systems resulting in an economic depression or worse.

If you cannot think of any, please explain why such a calamity is impossible in your opinion.

Roger

-- Dr. Roger Altman (rogaltman@aol.com), September 07, 1999.



Dave and Clyde:

As I read it, in making the same point, the two of you make the same mistake Yourdon does. Both of you assume that y2k is all-or-nothing, basically. You are saying that remaining bugs (after remediation and testing, in most cases) are still fatally massive, and that there's no fallback position. Once again, the implication is that either the software status at rollover is essentially perfect, or is hugely disastrous. Nothing in between, despite all the effort put into remediation, all the testing that's been done, or any workaround that human ingenuity can dream up to create a suboptimal but still workable system.

From reading the endgame essay and your responses, I tend to conclude that the entire concept of "suboptimal" doesn't exist in your universes (or in Yourdon's either). I've tried before to emphasize that our language recognizes that we live in an imperfect world (note the word "imperfect" here?). We also have words like "close", and "adequate", and "sufficient", and "mediocre".

You (and dit) seem to assume that this not-so-hypothetical company that's done a lousy job with all software they've touched, is running nonetheless on good software! Yeah, sure, they had to kill some turkeys, and some new things bombed and they had to go back to the prior stuff, but you assume the prior stuff is just fine. Why? If they're that lousy, the prior stuff is probably lousy too (in real life, it's pretty bad, it breaks a lot, the programmers are new and the code base is undocumented spaghetti. Fixes are slow, and introduce as many new problems as they cure old ones.)

But I can see that it's hard to communicate what I consider fairly obvious -- that Yourdon has built a case that *relies* on oversimplification, rather than simplified the real world for clarity. In the real world, a significant number of shops are *constantly* in the position of White in Figure 1! But the game goes on forever, lousy code layered onto lousy code, and ported from one platform to another whole cloth (including all the dead code, because you can no longer tell that it's dead, etc.) The chess game metaphor just doesn't describe reality very well, and Yourdon is building his argument on those elements of chess that reflect the real world the least.

So I can see it will be even harder to communication what I consider more subtle -- the (apparently) inevitable exaggeration of the importance of IT, on the part of those who make their living from it. Certainly I'm not trying to say that computers are NOT important, especially to large companies. They're critical, no doubt about it. But to hear longtime IT people like Yourdon and Hamasaki talk, every company that does anything useful revolves around IT like a planet around the sun. The corollary is that if IT sneezes, the company dies of pneumonia.

IT people are convinced that the company lives or dies based on software, much like the financial people think the company's lifeblood is money, upper management thinks it's leadership, and the union thinks it's labor. And those whose fear of y2k is most comprehensive seem to believe that if software fails, the money and labor will both vanish! (and the leadership was never there in the first place, I suppose).

And again (to coprolith here), though software is critical, date mishandling impacts don't have the sheer scope to knock the entire IT department down and keep it down, unable to accomplish a thing. Now, if assignment statements were to stop working on some known date... [grin]

While I fully realize my perspective is not well received around here, I still consider it self-evident that software after rollover will be even worse than usual. Still, I accept Hoffmeister's contention that the problems we're encountering implementing all these new systems (which are substantial, and well documented on this forum) are at least as serious and disruptive as the actual date bugs will be, that strike after rollover. And we're getting by with hardly a ripple. Not even the stock market has noticed yet.

I know Yourdon isn't happy that he said all controversy would end last April 1. I don't expect him to be any happier over the course of the next 18 months, as optimists keep asking him, "Hey Ed, when do we "lose" this chess game anyway?" But I do notice (with some amusement) that Yourdon was very careful neither to define "losing" nor to put a date on the "loss". Pessimists are increasingly careful to leave weasel room, for good reason.

***footnote to 'a': complex is not a synonym for "bad", as you imply. Our best econometric models are not very good even running on supercomputers -- the economy is too complex. But that hasn't meant the economy is bad by any definition. And y2k is much the same -- extremely complex, with millions of actors doing millions of different things simultaneously, many of which will interact with one another (pretty much the definition of an economy, eh?). It's always been this way, yet the results have been extraordinarily good. We call them "civilization". (And yes, I know, it's been going downhill since the days of Pliny the Elder, a process that accelerated after you passed puberty).

Mr. Altman:

I'll give your request some thought. I'm not a big believer in dominoes, and your question seems to assume dominoes. But I did give one answer in my reponse here -- the sudden failure of assignment statements (in all languages)! THAT would do the trick, I think.

-- Flint (flintc@mindspring.com), September 07, 1999.


It sucks to have critics in your own backyard.

-- Stan Faryna (info@giglobal.com), September 07, 1999.

"What's interesting is, companies like this actually exist! And Yourdon knows it. Just for fun, let's turn this same question around a bit. If a company has screwed up every software project they've touched for 20 years and are still successful, and y2k is just one more software project to screw up, what is the likelihood that this company will soldier on just fine after screwing up one more? According to the metrics Yourdon is so fond of citing, most big software projects "fail", yet the companies suffering this fate continue on just fine."

I went as far as this paragraph and knew instantly that it was Flint, without having to look at the sig.

Y2K is not just any project they can screw up like the others and survive, let alone thrive. One Immovable Deadline for Everybody in the World at Once, Gazillions of Companies Messing With Their Systems at the Same Time. You see a difference Flint? (I'm sure someone already mentioned that, but I can't bother reading the rest of this thread, it's just another FlintFolly "look! I can type").

I wish the name of the thread-originator would show in "recent answers" also. If I'm right to think you're going to get worse as the weeks tick by (anxiety affect people differently), you'll be louder and grasping at thin air more often.

-- (not@now.com), September 07, 1999.


Flint,

Are you a beer drinker? What kind? Are you drinking some right now?

Are you a smoker? What kind? Are you smoking some right now?

I don't think Dr.Ed is opinionated and shrill. Or did you mean shill?

Regarding that last statement in your opening post, critical eye, simplistic assumptions. Is that how you read these Y2k reports, data, metrics? If it is, then no wonder you can't see the complex connections. Don't like dominoes? Why not? Domino theory and Chaos theory are "good stuff", really!

-- Gordon (gpconnolly@aol.com), September 07, 1999.


BZZZZT. Wrong, Flint. Yourdon (and myself, Cory, a and the rest) have been working with orgs for 20 years and more, some of which are awful but yes, indeedy, stay in business. Think IRS. Bull Systems was my premier personal experience. It was more amazing that they survived than that they didn't. But they did. You're loony if you think we don't recognize that phenomenon or that you have somehow "discovered it."

Or that it has jack to do with Yourdon's essay.

-- BigDog (BigDog@duffer.com), September 07, 1999.



Gordon:

If you agree with Yourdon's opinions, fine, we can debate and discuss and communicate. If you don't think Yourdon is opinionated, we cannot communicate at all. He is highly opinionated, however passionately you may agree with him. I may sincerely believe my own opinions reflect reality better than anyone else's (if there were a better opinion than mine, of course I'd hold it [grin]). But I certainly recognize that I'm very opinionated, however correct my opinions happen to be.

Big Dog:

I must have really stepped on your tail, for you to go into attack mode armed with so little forethought. If you cannot see any relationship between getting by for a long time with terrible software (IRS, Bull, etc.) and Yourdon's essay, what can I possibly tell you? Yourdon has used chess analogies to demonstrate just why IRS and Bull are "losers". And you point out that they both continue to "win" anyway, by staying around and accomplishing their goals. Amazing what you can do even with awful software, isn't it? That's just what I said. I swear if I wrote my name on your foot, you'd shoot it.

The argument, all in all, seems to be that y2k will cause too many failures to deal with, all in too short a time. And I believe that will be true here and there. But not so bad most of us will need our preparations to survive. Check back in a year, I guess...

-- Flint (flintc@mindspring.com), September 07, 1999.


Flint: As I reported in a post a couple of days ago, the manager of my bank does not think that bad data can be screened out of his bank's computer systems, despite the "significant testing" that you claim has ruled out the bad data bogeyman. Should the people on this forum believe what you say or what my banker says? Which is a more prudent choice?

-- cody (cody@y2ksurvive.com), September 07, 1999.

Flint,

It's always humbling to see someone completely miss the point of something I was attempting to communicate.  I'm sure the problem is not with you, but with me ... the point of my essay was not "winning" or "losing," but the "end-game" that many of us feel we're moving into.  We're down to only 115 days now, and time is growing short; the question is whether the outcome is still in question -- or whether the Y2K situation that we see today is essentially the same that we'll see on December 31st.& nbsp; It's like the situation when you take your kid to a baseball game, and your team is behind by 37 points at the top of the 9th; it's late at night, and you're tired, and all you want to do is go home.  But your kid wants to stay until the bitter end, because there's always a chance of a last-minute miracle.  Yeah, maybe so -- but it's a lot easier to believe if your team is down by only a run or two, and the top of the batting order is coming up during the final inning.

Well, since I didn't do so well with my chess metaphor, perhaps I'd better stop before I get too deeply involved in a baseball metaphor.& nbsp; Instead, let me respond to the points you've raised....

Yourdon's essay, if nothing else, indicates that he's becoming as opinionated and shrill as I am.
The Microsoft Bookshelf dictionary defines "opinionated" as "Holding stubbornly and often unreasonably to one's own opinions."  Well, if that's < i>your opinion, I respect it even though I disagree.  I don't know how carefully you're using your words, by the way; for example, the same dictionary offers several definitions of "opinion," one of which is "A belief or conclusion held with confidence but not substantiated by positive knowledge or proof." and another of which is "A judgment based on special knowledge and given by an expert: a medical opinion." I'd like to think that my 35 years of experience in the computer field, and my years of writing about, and consulting on, the management of large, complex software projects qualifies me for the latter definition...
And as a result, he seems to be wandering off into the deep end without realizing it.
The "deep end" of what?  Well, who cares ... I won't pursue this one any further.
The whole essay uses a chess game as a metaphor. This is very instructive, because what's critical is that a chess game has a defined, clear winner and loser.
Perhaps that's true where you come from.  Where I come from, a game of chess can end in a win, loss, or draw.  I apologize for not making that point clear for those unaware of the game.
And this isn't a trivial observation either, because Yourdon relies throughout the essay on winning and losing at the y2k game, as though they were as clearly defined as in chess. Yourdon is more subtlethan most authors, but in the end, the implication is that your bank, your town, your utility either "wins" because its y2k work is utterly completed, or it "loses" by becoming entirely nonfunctional. No bank, no power, no police or firemen.
I wasn't trying to be subtle, but if I created a subtle misunderstanding, I apologize.  I concede your point that a novice might read the essay and conclude that the outcome of Y2K must be a complete win, or a complete loss.  But anyone who has been half-alert during four years of discussions about Y2K knows that there is a wide range of possible scenarios between "non-event" and TEOTWAWKI.  I've written about this until I'm blue in the face, and I've used the Washington DC Y2K User Group's "scale" as a reasonable way of discussing this.  It seems to me that if a Y2K problem causes a disruption in the supply of some important product or service for a week, that's worse than a disruption of only a few hours; if the disruption lasts for a month, that's a heck of a lot worse than a week.  The Montreal ice storm of 1998, for example, would not be characterized as a complete "loss" for the utility companies -- people eventually got their power restored.  But during the period of nearly a month that many of those residents had no electric power, I doubt that they would have been pleased to hear the local power company bragging about how they "won" the battle against Mother Nature.
Once this metaphor is established, Yourdon uses it to expound on his favorite topic -- how software projects don't come in on time, how nobody meets his satisfaction with their current status, and how many significant economic actors are hopelessly behind and will "lose" the endgame. If you buy into the all-or-nothing nature of the chess metaphor, it all seems to add up.
Argh!  The point of the essay was to expound upon the end game, not my "favorite topic" of projects finishing late.  And something else that I obviously failed to communicate is that every individual can form a different opinion of the status of the end-game.  I'll certainly admit that I have a more jaundiced, pessimistic outlook on the situation than many people do, because of the many software projects I've seen.& nbsp; But you're perfectly welcome to use the kind of logic that you expand upon later in your commentary to justify a more hopeful assessment of the end-game situation.
But buried in the middle of this long essay is an interesting sentence. Yourdon is talking about the track records of the actors, and rhetorically asks how much confidence you'd have in a company that had screwed up every software project they'd touched for 20 years. Well, I'm not in a position to tell what's "right" for you.  What I was trying to suggest is that if a company has screwed up every project they've touched for the past 20 years, there's a reasonably good chance they'll screw up the next one.  Predicting the outcome of a software project is not like predicting the outcome of a toss of a coin, where each toss is (in theory, at least) independent of all the previous tosses.& nbsp; If you're a baseball player who has struck out every time for the last 3,000 at-bats, how much confidence do you have that he'll hit a grand slam home run the next time he comes to the plate?
 The "right" answer, of course, is that this company will "lose".   What's interesting is, companies like this actually exist!
I disagree.  The baseball player who strikes out 3,000 times in a row is not necessarily a loser; he may continue to be employed by the same team, year after year.  Maybe it's because he's a good pitcher ... or maybe he's been able to get away with it, even though it's the worst record in history.  Maybe he married the daughter of the team owner.  In the "real world" of companies, government agencies, and software projects, there are companies that screw up projects time after time, but manage to avoid losing completely, in the sense that they don't go bankrupt.  Government agencies dobn't have to make a profit, so (in the extreme case) they can continue to screw up projects and waste money for decades without anything changing.  Companies can screw up projects and survive (even prosper!) because they own a key patent, or because they enjoy a monopoly in their industry, or because the CEO married the President's daughter ... or because the wheels of justice grind exceedingly slow, and decades pass before the inefficient, obsolete monolith finally crumbles.
And Yourdon knows it.
I'm always amazed by statements like this, made by people who have never met met me.  My parents, my wife, my siblings, and my children would be cautious about such a statement.  When a complete stranger makes such a comment, I have to wonder whether he uses the same logic to predict the outcome of software projects and Y2K initiatives in companies he has never visited, run by managers he's never met.
Just for fun, let's turn this same question around a bit. If a company has screwed up every software project they've touched for 20 years and are still successful, and y2k is just one more software project to screw up, what is the likelihood that this company will soldier on just fine after screwing up one more? According to the metrics Yourdon is so fond of citing, most big software projects "fail", yet the companies suffering this fate continue on just fine.
Perhaps this is the crux of our disagreement.  Sure, if Y2K is just one more project to screw up, then why are we making such a big fuss?& nbsp; But I hope it won't be too much of a surprise to you to learn that Y2K projects have a much more non-negotiable deadline than typical projects.  Typical projects can be a day late, a week late, six months late, or even two years late without necessarily being classified as a total failure.  Sure, the end-user may be annoyed, and the company may have missed its "window of opportunity" for capturing a new market -- but up to a point, the old saying "better late than never" is valid.  Even with Y2K projects, that may be true in a limited sense -- e.g., we don't have to get the financial systems completely done until January 31st when we try to close the books, and we don't have to get all the leap-year bugs fixed until February 29, 2000.  But the degree of flexibility is much, much less.
Maybe a better metaphor for remediation activities would be trying to negotiate a big raise from your boss. You ask for 20%, and he counters with 2%. If you end up agreeing on, say, 10%, did you "win" or "lose" this game?
Ummm... I don't know.  It depends on how desperate you were to get the full 20%.  Obviously, I agree that in most real-world situations, 10% is better than nothing at all, and we could probably find a way to muddle through with "only" 10% more income than we had before.  But if you have to prove that your income is 20% higher than today's level in order to qualify for a mortgage or to be approved for membership in the local golf club, then 10% won't do it.  I think we would agree that there are some situations where it's "all or nothing," and other situations where the degree of winning or losing is a relative thing.
I don't consider this an idle question at all. If your bank sends you a statement with the year as 1900 but the amount correct, did they "lose" the chess game? If your electric utility skips a few bills and then hits you with one big one for 3 months' worth, but the power was up all that time, did your utility "lose"? If every stoplight in town goes to blinking red for a day (with big lines everywhere) before it's corrected, did your town "lose"? If some of GM's assembly lines are down for a day, or WalMart can't restock some orders for a week longer than usual, or gasoline prices rise 50 cents a gallon for 6 months before dropping back down, are these all "losses"?
These are all "relative" losses.  Most of the ones you've constructed are obviously very minor losses ... but I'd be a little careful before casually suggesting a 50 cent rise in gasoline prices for six months.& nbsp; Even if gas prices returned to normal, that six-month "spike" could be enough to trigger a nasty economic recession.  And as you can imagine, I can just as easily concoct a series of examples in which the Y2K disruptions are far more significant.
Read what Yourdon says about banking. He can find only two problems -- that bad data may spread corruption, and that there might be cross defaults from "noncompliant"
customers.
Well, I'll accept the criticism that I provided only a superficial description of several possible examples of Y2K end-game situations.  The point of incorporating this into my essay was not to "prove" the outcome of the banking situation, one way or the other, but to provide some illustrations of how I've categorized the "end game" situation for myself in a few key areas.  It's a bit glib for you to say that I "can only find two problems"; I was merely illustrating two of the problems that make me conclude "the jury is out" when it comes to the banking situation.
But in real life, a significant level of testing has effectively ruled out the "bad data" boogyman,
Maybe that's true in your life, and maybe your life is more real than mine ... I'm willing to concede the strong likelihood that most large banks have done a lot of testing of their interfaces.  Whether "quantity" of  testing is the same as "quality" of testing is a matter of opinion, if for no other reason that we have no independent IV&V results to look at.  I have no idea what kind of test coverage has been provided, nor am I aware of any published results showing that all of the major banks around the world have done a sufficient level of testing that a majority of independent, competent computer professionals would agree that the bad data boogyman has been "effectively ruled out."  The reports that I've seen from S.W.I.F.T. seem to indicate a fair amount of frustration about the low degree of participation by global banks in their testing efforts.  I'd like to be hopeful about the banking situation, and I think I am much more hopeful than, say, Gary North, who seems to feel that the "game" involving bad data corruption has been irrevocably lost.  For me, the jury is still out; we're in the end game, and I'll keep watching the situation every day.  If you've decided -- based on your own knowledge and experience -- that this problem is categorically, unquestionably solved, that's terrific.  You'll sleep easily for the next 115 days, and I won't.
and noncompliance does not equate with death (or inability to repay a loan) any more than a 20 year history of software incompetence has meant this in the past.
Huh?  That's a little glib, isn't it?  If noncompliance create any greater risk of loan default than a 20 year history of software incompetence, then why are banks putting so much pressure on their large customers to demonstrate that they will be compliant?  For that matter, why should any company bother asking any of its suppliers or vendors for reassurances about compliance?  Why has the SEC told brokerage firms they'll be shut down if they can't demonstrate compliance by whatever deadline they're currently using?  Let's not be silly about this: obviously, noncompliance does not guarantee death or bankruptcy or inability to repay a loan.  But when I read survey after survey indicating that 40-50% of small businesses don't plan to remediate any of their systems, I have to assume it could cause some problems for their mission-critical activities in addition to the minor systems that maintain the inventory of toilet-paper stock.  We've also got surveys from organizations like Cap Gemini indicating that a modest percentage of Fortune 500 companies will be left with as many as 25% of their mission-critical systems unremediated ... and you're suggesting to me that I don't have to worry at all about the possibility of this causing some significant loan defaults?  You manage to sleep well at night with the prospect of entire countries -- Brazil, Indonesia, Russia being the first ones that come to mind -- defaulting on their loans?  Well, more power to you ... in my little head, it just doesn't compute very well.
So Yourdon is building his case quite clearly on policy statements both contrary to observation.  But he's careful not to mention this -- apparently it undermines his black-and-white world, for purposes of this essay.


Huh? Policy statements?  Whose policy statements?

Anyway, what I'm trying to say is that Yourdon's fundamental assumption, that y2k remediation is all-or-nothing, and all the players either achieve perfection or "lose" and die, is incorrect. He's subtle about it, but still wrong. Instead of reading this essay with the TB2000 Will To Believe, try reading it with a critical eye to simplistic assumptions. And it doesn't hold water.
As I've tried to explain, I do not believe that Y2K remediation is all-or-nothing, but to a much larger extent than with any traditional projects, the deadline aspect of project success/failure is indeed all-or-nothing.  As for the binary outcomes of "perfection" versus "lose," that too is a red herring in your argument.  As a reasonably intelligent observer of, say, the banking situation, I know that there is a wide range of possible outcomes: perfection (nothing at all goes wrong); minor inconvenience (my ATM is down for a couple days, but is then restored); moderate inconvenience (the bank is closed for a week, and all ATM's in the region are down for a month); major inconvenience (my bank fails, and it takes the FDIC a year to repay my insured deposits); and disaster (the entire banking system fails).

The real question I was trying to address in my essay is the end- game question: given that it's now September 7th, should I assume that the banking situation is likely to change dramatically for the better in the next 115 days? Or change dramatically for the worse in the next 115 days?  Or is it all over but the shouting?  Should I wait for another week to see if I can get a clearer picture of the situation?  Another month?  Two months?  At what point have I waited too long to exercise the options that I could have exercised today?  None of this may be of any concern to you, Flint, but it's something that I wrestle with every day -- and I think many others, do, too.

-- Ed Yourdon (HumptyDumptyY2K@yourdon.com), September 07, 1999.


another attempt ... hopefully with a couple of HTML glitches removed...

Flint,

It's always humbling to see someone completely miss the point of something I was attempting to communicate. I'm sure the problem is not with you, but with me ... the point of my essay was not "winning" or "losing," but the "end-game" that many of us feel we're moving into. We're down to only 115 days now, and time is growing short; the question is whether the outcome is still in question -- or whether the Y2K situation that we see today is essentially the same that we'll see on December 31st. It's like the situation when you take your kid to a baseball game, and your team is behind by 37 points at the top of the 9th; it's late at night, and you're tired, and all you want to do is go home. But your kid wants to stay until the bitter end, because there's always a chance of a last-minute miracle. Yeah, maybe so -- but it's a lot easier to believe if your team is down by only a run or two, and the top of the batting order is coming up during the final inning.

Well, since I didn't do so well with my chess metaphor, perhaps I'd better stop before I get too deeply involved in a baseball metaphor. Instead, let me respond to the points you've raised....

Yourdon's essay, if nothing else, indicates that he's becoming as opinionated and shrill as I am.
The Microsoft Bookshelf dictionary defines "opinionated" as "Holding stubbornly and often unreasonably to one's own opinions." Well, if that's your opinion, I respect it even though I disagree. I don't know how carefully you're using your words, by the way; for example, the same dictionary offers several definitions of "opinion," one of which is "A belief or conclusion held with confidence but not substantiated by positive knowledge or proof." and another of which is "A judgment based on special knowledge and given by an expert: a medical opinion." I'd like to think that my 35 years of experience in the computer field, and my years of writing about, and consulting on, the management of large, complex software projects qualifies me for the latter definition...
And as a result, he seems to be wandering off into the deep end without realizing it.
The "deep end" of what? Well, who cares ... I won't pursue this one any further.
The whole essay uses a chess game as a metaphor. This is very instructive, because what's critical is that a chess game has a defined, clear winner and loser.
Perhaps that's true where you come from. Where I come from, a game of chess can end in a win, loss, or draw. I apologize for not making that point clear for those unaware of the game.
And this isn't a trivial observation either, because Yourdon relies throughout the essay on winning and losing at the y2k game, as though they were as clearly defined as in chess. Yourdon is more subtlethan most authors, but in the end, the implication is that your bank, your town, your utility either "wins" because its y2k work is utterly completed, or it "loses" by becoming entirely nonfunctional. No bank, no power, no police or firemen.
I wasn't trying to be subtle, but if I created a subtle misunderstanding, I apologize. I concede your point that a novice might read the essay and conclude that the outcome of Y2K must be a complete win, or a complete loss. But anyone who has been half-alert during four years of discussions about Y2K knows that there is a wide range of possible scenarios between "non-event" and TEOTWAWKI. I've written about this until I'm blue in the face, and I've used the Washington DC Y2K User Group's "scale" as a reasonable way of discussing this. It seems to me that if a Y2K problem causes a disruption in the supply of some important product or service for a week, that's worse than a disruption of only a few hours; if the disruption lasts for a month, that's a heck of a lot worse than a week. The Montreal ice storm of 1998, for example, would not be characterized as a complete "loss" for the utility companies -- people eventually got their power restored. But during the period of nearly a month that many of those residents had no electric power, I doubt that they would have been pleased to hear the local power company bragging about how they "won" the battle against Mother Nature.
Once this metaphor is established, Yourdon uses it to expound on his favorite topic -- how software projects don't come in on time, how nobody meets his satisfaction with their current status, and how many significant economic actors are hopelessly behind and will "lose" the endgame. If you buy into the all-or-nothing nature of the chess metaphor, it all seems to add up.
Argh! The point of the essay was to expound upon the end game, not my "favorite topic" of projects finishing late. And something else that I obviously failed to communicate is that every individual can form a different opinion of the status of the end-game. I'll certainly admit that I have a more jaundiced, pessimistic outlook on the situation than many people do, because of the many software projects I've seen. But you're perfectly welcome to use the kind of logic that you expand upon later in your commentary to justify a more hopeful assessment of the end-game situation.
But buried in the middle of this long essay is an interesting sentence. Yourdon is talking about the track records of the actors, and rhetorically asks how much confidence you'd have in a company that had screwed up every software project they'd touched for 20 years.

Well, I'm not in a position to tell what's "right" for you. What I was trying to suggest is that if a company has screwed up every project they've touched for the past 20 years, there's a reasonably good chance they'll screw up the next one. Predicting the outcome of a software project is not like predicting the outcome of a toss of a coin, where each toss is (in theory, at least) independent of all the previous tosses. If you're a baseball player who has struck out every time for the last 3,000 at-bats, how much confidence do you have that he'll hit a grand slam home run the next time he comes to the plate?

The "right" answer, of course, is that this company will "lose". What's interesting is, companies like this actually exist!
I disagree. The baseball player who strikes out 3,000 times in a row is not necessarily a loser; he may continue to be employed by the same team, year after year. Maybe it's because he's a good pitcher ... or maybe he's been able to get away with it, even though it's the worst record in history. Maybe he married the daughter of the team owner. In the "real world" of companies, government agencies, and software projects, there are companies that screw up projects time after time, but manage to avoid losing completely, in the sense that they don't go bankrupt. Government agencies dobn't have to make a profit, so (in the extreme case) they can continue to screw up projects and waste money for decades without anything changing. Companies can screw up projects and survive (even prosper!) because they own a key patent, or because they enjoy a monopoly in their industry, or because the CEO married the President's daughter ... or because the wheels of justice grind exceedingly slow, and decades pass before the inefficient, obsolete monolith finally crumbles.
And Yourdon knows it.
I'm always amazed by statements like this, made by people who have never met me. My parents, my wife, my siblings, and my children would be cautious about such a statement. When a complete stranger makes such a comment, I have to wonder whether he uses the same logic to predict the outcome of software projects and Y2K initiatives in companies he has never visited, run by managers he's never met.
Just for fun, let's turn this same question around a bit. If a company has screwed up every software project they've touched for 20 years and are still successful, and y2k is just one more software project to screw up, what is the likelihood that this company will soldier on just fine after screwing up one more? According to the metrics Yourdon is so fond of citing, most big software projects "fail", yet the companies suffering this fate continue on just fine.
Perhaps this is the crux of our disagreement. Sure, if Y2K is just one more project to screw up, then why are we making such a big fuss? But I hope it won't be too much of a surprise to you to learn that Y2K projects have a much more non-negotiable deadline than typical projects. Typical projects can be a day late, a week late, six months late, or even two years late without necessarily being classified as a total failure. Sure, the end-user may be annoyed, and the company may have missed its "window of opportunity" for capturing a new market -- but up to a point, the old saying "better late than never" is valid. Even with Y2K projects, that may be true in a limited sense -- e.g., we don't have to get the financial systems completely done until January 31st when we try to close the books, and we don't have to get all the leap-year bugs fixed until February 29, 2000. But the degree of flexibility is much, much less.
Maybe a better metaphor for remediation activities would be trying to negotiate a big raise from your boss. You ask for 20%, and he counters with 2%. If you end up agreeing on, say, 10%, did you "win" or "lose" this game?
Ummm... I don't know. It depends on how desperate you were to get the full 20%. Obviously, I agree that in most real-world situations, 10% is better than nothing at all, and we could probably find a way to muddle through with "only" 10% more income than we had before. But if you have to prove that your income is 20% higher than today's level in order to qualify for a mortgage or to be approved for membership in the local golf club, then 10% won't do it. I think we would agree that there are some situations where it's "all or nothing," and other situations where the degree of winning or losing is a relative thing.
I don't consider this an idle question at all. If your bank sends you a statement with the year as 1900 but the amount correct, did they "lose" the chess game? If your electric utility skips a few bills and then hits you with one big one for 3 months' worth, but the power was up all that time, did your utility "lose"? If every stoplight in town goes to blinking red for a day (with big lines everywhere) before it's corrected, did your town "lose"? If some of GM's assembly lines are down for a day, or WalMart can't restock some orders for a week longer than usual, or gasoline prices rise 50 cents a gallon for 6 months before dropping back down, are these all "losses"?
These are all "relative" losses. Most of the ones you've constructed are obviously very minor losses ... but I'd be a little careful before casually suggesting a 50 cent rise in gasoline prices for six months. Even if gas prices returned to normal, that six-month "spike" could be enough to trigger a nasty economic recession. And as you can imagine, I can just as easily concoct a series of examples in which the Y2K disruptions are far more significant.
Read what Yourdon says about banking. He can find only two problems -- that bad data may spread corruption, and that there might be cross defaults from "noncompliant"
customers.
Well, I'll accept the criticism that I provided only a superficial description of several possible examples of Y2K end-game situations. The point of incorporating this into my essay was not to "prove" the outcome of the banking situation, one way or the other, but to provide some illustrations of how I've categorized the "end game" situation for myself in a few key areas. It's a bit glib for you to say that I "can only find two problems"; I was merely illustrating two of the problems that make me conclude "the jury is out" when it comes to the banking situation.
But in real life, a significant level of testing has effectively ruled out the "bad data" boogyman,
Maybe that's true in your life, and maybe your life is more real than mine ... I'm willing to concede the strong likelihood that most large banks have done a lot of testing of their interfaces. Whether "quantity" of testing is the same as "quality" of testing is a matter of opinion, if for no other reason that we have no independent IV&V results to look at. I have no idea what kind of test coverage has been provided, nor am I aware of any published results showing that all of the major banks around the world have done a sufficient level of testing that a majority of independent, competent computer professionals would agree that the bad data boogyman has been "effectively ruled out." The reports that I've seen from S.W.I.F.T. seem to indicate a fair amount of frustration about the low degree of participation by global banks in their testing efforts. I'd like to be hopeful about the banking situation, and I think I am much more hopeful than, say, Gary North, who seems to feel that the "game" involving bad data corruption has been irrevocably lost. For me, the jury is still out; we're in the end game, and I'll keep watching the situation every day. If you've decided -- based on your own knowledge and experience -- that this problem is categorically, unquestionably solved, that's terrific. You'll sleep easily for the next 115 days, and I won't.
and noncompliance does not equate with death (or inability to repay a loan) any more than a 20 year history of software incompetence has meant this in the past.
Huh? That's a little glib, isn't it? If noncompliance create any greater risk of loan default than a 20 year history of software incompetence, then why are banks putting so much pressure on their large customers to demonstrate that they will be compliant? For that matter, why should any company bother asking any of its suppliers or vendors for reassurances about compliance? Why has the SEC told brokerage firms they'll be shut down if they can't demonstrate compliance by whatever deadline they're currently using? Let's not be silly about this: obviously, noncompliance does not guarantee death or bankruptcy or inability to repay a loan. But when I read survey after survey indicating that 40-50% of small businesses don't plan to remediate any of their systems, I have to assume it could cause some problems for their mission-critical activities in addition to the minor systems that maintain the inventory of toilet-paper stock. We've also got surveys from organizations like Cap Gemini indicating that a modest percentage of Fortune 500 companies will be left with as many as 25% of their mission-critical systems unremediated ... and you're suggesting to me that I don't have to worry at all about the possibility of this causing some significant loan defaults? You manage to sleep well at night with the prospect of entire countries -- Brazil, Indonesia, Russia being the first ones that come to mind -- defaulting on their loans? Well, more power to you ... in my little head, it just doesn't compute very well.
So Yourdon is building his case quite clearly on policy statements both contrary to observation. But he's careful not to mention this -- apparently it undermines his black-and-white world, for purposes of this essay.


Huh? Policy statements? Whose policy statements?

Anyway, what I'm trying to say is that Yourdon's fundamental assumption, that y2k remediation is all-or-nothing, and all the players either achieve perfection or "lose" and die, is incorrect. He's subtle about it, but still wrong. Instead of reading this essay with the TB2000 Will To Believe, try reading it with a critical eye to simplistic assumptions. And it doesn't hold water.
As I've tried to explain, I do not believe that Y2K remediation is all-or-nothing, but to a much larger extent than with any traditional projects, the deadline aspect of project success/failure is indeed all-or-nothing. As for the binary outcomes of "perfection" versus "lose," that too is a red herring in your argument. As a reasonably intelligent observer of, say, the banking situation, I know that there is a wide range of possible outcomes: perfection (nothing at all goes wrong); minor inconvenience (my ATM is down for a couple days, but is then restored); moderate inconvenience (the bank is closed for a week, and all ATM's in the region are down for a month); major inconvenience (my bank fails, and it takes the FDIC a year to repay my insured deposits); and disaster (the entire banking system fails).

The real question I was trying to address in my essay is the end- game question: given that it's now September 7th, should I assume that the banking situation is likely to change dramatically for the better in the next 115 days? Or change dramatically for the worse in the next 115 days? Or is it all over but the shouting? Should I wait for another week to see if I can get a clearer picture of the situation? Another month? Two months? At what point have I waited too long to exercise the options that I could have exercised today? None of this may be of any concern to you, Flint, but it's something that I wrestle with every day -- and I think many others, do, too.

-- Ed Yourdon (HumptyDumptyY2K@yourdon.com), September 07, 1999.


Well, it looked fine when I typed it and displayed it with Netscape Composer ... but it looks like a couple of extra indentations (blockquotes) snuck in there.

Sysop(s), could you delete the first version of my posting to avoid clutter? And if necessary, could you stick in a couple of right- indents or whatever is required to keep subsequent postings from dribbling off the right side of the screen?

Muchos gracias

Ed

-- Ed Yourdon (HumptyDumptyY2K@yourdon.com), September 07, 1999.


Ed:

That's why I never try to use HTML. I doubt anything I post would be readable at all if I tried. Or maybe I should say legible?

Anyway, if I misread what you were driving at, I apologize, although I still think we disagree about the eventual impacts that date bugs will have (and I couldn't begin to guess what fear of date bugs, in and of itself, might cause).

No, I don't think a great deal more will be accomplished between now and the end of the year, at least not at a rate greater than has been happening for the last year or so. For many, that rate represents flat out, faster than which they cannot go (and make things worse by trying). And as to how much actually has been accomplished, I don't claim any special insights either.

But let's look at the bank testing as an illustration of divergent viewpoints. I concede that international bank testing has not attracted as many banks as everyone would prefer. I'm encouraged that the major banks have tested, and also encouraged that the testing apparently went fairly well.

Now, you could argue that (1) We really don't know how well the testing went, we have only their PR to go by; (2) In any case, only the most-prepared banks engaged in this testing; and (3) The testing may have failed to cover what will turn out to be critical -- test coverage of code is never complete.

But then, where does the truth lie? I take it this is your point. It *might* be that not enough banks have tested, with tests we don't know enough about, with results we aren't privy to. But (if I read you right), by the time (if it ever comes at all) when we have enough information, we won't have sufficient time to act on it if it's bad news.

This point is well taken. I feel you are exaggerating the danger based on my reading of everything I can get my hands on, and based on my own experience. The trend I see forming right now is that things aren't quite as bad as feared, tests are more common and going better than expected, major players are putting the brakes on their remediation projects, and I feel you are exaggerating the status of small businesses (I wrote at more length on this on a recent thread about Australian businesses).

I'd say right now I think the world is humming along at about an A- but y2k will reduce this to B- for a while. Collapse would require an F, and I sincerely doubt we'll reach that point, barring total public panic of some form. I see failure as being a substantial distance south of adequate, and I don't view a much higher but still manageable error rate as failure. But I admit it's hard to draw contrasts for purposes of illustration without creating falsehoods, in a world of infinite gradations.

-- Flint (flintc@mindspring.com), September 07, 1999.


Flint,

OK, it comes down to this. What does opinionated mean? Sure, Ed has stong opinions, but he is also very careful to constantly remind his readers that he is stating his own opinion, and that you need to find *your* own opinion. Is that opinionated? Because if it is, then everyone is opinionated if they hold any opinion on anything. Just for the record, I don't consider you opinionated either, sorry. I can't charge you with that. Debate-ated yes, that you are. But I've gotten used to that part and don't mind that either.

-- Gordon (gpconnolly@aol.com), September 07, 1999.


Oh, a PS to Ed:

If I remember correctly, you have talked about incompetent organizations (especially in government, but private sector too) in several of your essays and in some of your books as well. I felt safe in saying you knew such companies exist, since you've discussed them!

-- Flint (flintc@mindspring.com), September 07, 1999.


Flint, go ahead and smoke. PLEASE!

-- (enough@already.Flint), September 07, 1999.

Flint,

One more thought for you. You probably remember the alleged story of Harry Truman complaining about economists. "Give me a one-handed economist" he said. "Because whenever I ask them what will the economy be like next year, they say, on the one hand it could be this, and on the other hand, it could be that." You can relate to that, right?, because you do that sort of thing all the time. Ed, in contrast, was exactly what old Harry was looking for. Ed probably could have said something like: "Well, this is only my own opinion, you understand, but I think there will be a year of disruptions followed by a 10 year depression." And then he would have explained just why he thought it would turn out that way, with details. Is that *opinionated*, or merely willing to answer a question with a definite strong opinion?

-- Gordon (gpconnolly@aol.com), September 07, 1999.


Gordon:

A good point. I was contrasting Yourdon's current position with his TB2000 book, wherein each chapter ended with a list of alternatives depending on what level of event you expected or could guard against.

And I've tried to do this as well. To paraphrase Yourdon, if you've misunderstood that badly, it can only be that I expressed myself that poorly. I've tried to make it clear that I don't expect y2k to be a catastrophe. I expect plenty of problems, but relatively few to be beyond what we can manage and deal with (however imperfectly). I expect delays and inconveniences and short tempers and VERY long days for the geeks (possibly myself included, depending on what people decide to do with certain embedded systems). I do NOT expect riots, nor more than fairly short and local power outages, nor a collapse of the banking system, nor lack of dial tone over large parts of the world, nor more than a small reduction in imports (very much contrary to the Paul Milne NOT ONE DROP of imported oil, fershluggener!) I don't expect a return to the gold standard (I think gold will turn out to be a lousy play), nor a collapse of any Western government. In my post to start this thread, I gave a paragraph full of the kinds of impacts I expect to be most common (though a few real bad ones wouldn't surprise me either).

NOW, I try to justify my expectations as well as I can, given that my crystal ball is no better than yours. I make an attempt to show that I consider what I read at least somewhat carefully, and at least I'm aware of it. I try to show that there are many ways to look at things, and therefore informed conclusions are necessarily tentative conclusions. I've given my conclusions, but they ARE tentative -- new information continues to flow in. OK, Yourdon interprets what he reads as Figure 2 at best, while I read most of it as being like Black in Figure 1, and figure 2 as worst case (with some horrible exceptions, mostly in government).

Ed answered my post carefully (much more careful, indeed, than I was when I wrote it), and I appreciate this a lot.

-- Flint (flintc@mindspring.com), September 07, 1999.


Flint, the more I read your postings the more I become aware that the big picture has eluded you, either from a chronic lack of nicotine or because of your own intellectual dishonesty.

You claim that rioting will not be a result of the coming crisis, yet the second most populous country on Earth burns as we speak. You claim that "banks get it" yet instead of documented test results on the countless interfaces, all we get is an industry form letter that says withdrawing your money is stupid. You're probably one of those optimistic idiots that thinks the DOW will hit 36000 in a couple years. Flint, let me tell you something: if the world worked the way you have described, we would still be ruled by the Romans or Egyptians, because according to you, complex systems can't screw up.

When it comes time to personally assess y2k and its related effects, you are going to find that you have been a patsy, plain and simple.

BTW - Ed's rebuttal was on target. I haven't seen smoke like that coming from you since you kicked your two pack a day habit and stopped matching wits with Milne.

-- a (a@a.a), September 07, 1999.


Flint,

Yes, Ed is a gentleman, all the way. But so are you, for the most part, except when you get a little mad, which isn't very often. I realize you have a different outlook on what will happen, and you've made that very clear, many times. Difference of opinion is all. No problem there. I think I got a little disappointed by the use of just a couple of words, like opinionated and shrill. Otherwise, it doesn't bother me that you find Ed's approach in error. And I guess it doesn't bother Ed all that much either. I'm sure he's been challenged that way many times by many people in the last 18 months. We have both heard the suggestion "Let's just agree to disagree" many times. In the past, it has been *just* a suggestion, or concept, in my dealings with others. With you, though, it has become a working reality. I've learned something from your posts, your attitude, your entire approach to defining this slippery Y2k snake in the grass. In some strange way you have actually made me more aware of tolerance in handling the opinions of others that are in conflict with my own. Y2k is such a nebulous, theoretical predicament, that it has strained all of us in trying to get our arms around it. You've tried, I've tried, Ed's tried. All of us here in one form or another. It can reach right into our deepest fears and frustrations, that's for sure.

-- Gordon (gpconnolly@aol.com), September 07, 1999.


'a':

I'm prepared to be wrong, and willing to admit it if events prove out that way. Certainly I think there's enough ambiguity for me to have WAY misinterpreted the signs. You ought to admit the same. Time will tell, I hope.

You yourself are well aware that the publication of any details of banking interface testing would be a hacker's dream come true. And if instead of the detail we got summaries of those results, well, that's what we've got and you label them inadequate. Asking for the impossible and saying nothing short of that is sufficient is a straw man argument. It comes down to whether you trust the banks. On the one hand, they urgently need public confidence. On the other hand, they urgently need properly working software. Your choice, I guess. I trust them because by FIXING the damn problems, they can kill both birds with the same stone, except in the minds of the blind denialists.

I've never really argued that complex systems can't screw up. I've argued that they're resiliant, but you're right, sooner or later they all have died. But it's even LESS rational to argue that because it's *possible* for them to screw up, therefore (this time) they are *sure* to do so. Not all possibilities are guaranteed, quite the contrary.

As for Milne, I was certainly at a disadvantage, there being no wits there for me to try to match.

-- Flint (flintc@mindspring.com), September 07, 1999.


How many angles can fit on the head of a pin? Intelligent men debated this question for years, or so I have been told.

Time will tell who is right and who is wrong. Flint seems to be the ultimate polly. Arguing there will be no colapse, but preparing none the less. Sort of like a stewardess telling passangers "stay in your seats, stay calm, while she is putting on a parachute". Rational thought leads one to prepare for the worst and hope for the best. I respect and thank Mr. Yourdon for his thoughts and warnings on this unique computer problem.

-- Bill (y2khippo@yahoo.com), September 07, 1999.


Yes, thank you Mr. Yourdon.

On a thread I started about two months ago Flint stated that "Ed was his hero" now Ed's essay "doesn't hold water".

Tsk tsk, you are not a Jedi yet Flint.

-- Will (sibola@hotmail.com), September 08, 1999.


Ed - Flint

I want to thank you both for a truly wonderful discussion and debate. I am truly impressed by both of your opinions. This whole thread still hasn't changed my mind, though. We are prepped for a year + ...and are truly hoping that we won't need it.

Flint - I hope (and even somewhat suspect) that you are correct in most of your assessment of the y2k bug issue.

BUT - when you step back and take a look at the rest of the ingredients for the millenium cake we are baking (i.e. cyber terrorism, panic in the streets for general purpose reasons and the growing tensions around the globe) - I have come to the conclusion that 2000-2001 are going to be hard years and ones not easily put aside.

So does it really matter which straw breaks the camel's back? History will lump it all into one big (or small) paragraph anyhow.

Mr. Yourdon - I want to thank you for this wonderful forum you tossed together. I have been lurking for almost a year now - occasionally posting. I have so few original opinions that others can't express better (sigh). I liked your essay just fine. It clicked and made sense - I got it on the first try. Then flint made the comment about preaching to the choir - and, upon rereading the essay, I found that he is correct. DGIs and newcomers would have a tendency to think all or nothing. Yes Ed, there are still folks out there that know very little about the whole issue. But your essay should provide the impetus for further action on some of their parts.

IMHO - I don't think that either of you are too far apart on your positions.

thanks again. I apologize if I sound like I am rambling - insomnia at 2 am has a tendency to do that. Just know that I felt strong enough about this that I had to come out of lurker status to comment.

And it's okay to flame if you have to...my housecoat is flame- retardent.

-- justme (finally@home.com), September 08, 1999.


Just me,

Yes, I agree 100%. Y2k won't happen in a vacuum. It is a huge jolt of uncertainty, a splash of ice cold water below the waist. It is a time of collective introversion and reflection (as in "how the heck did we get where we are now and did we really do the right thing"), not exhuberant extroversion and optimism. Because the economy of today not only thrives on but relies on confidence and risk-taking (and tolerates nothing less), recession is inevitable in 00. Even with NO computer errors, you can imagine a situation where people, industries, and countries prepare for this Phantom Menace or exploit it to their advantage:

(1) Oil companies, anticipating y2k problems and fuel stockpiling, stick it to the customer and gouge him with 3 bucks per gallon of gas. This makes up for the lack of profits suffered by the industry since the mid 80s.

(2) Countries alleged to have bad y2k problems default on all their loans and/or devalue their currencies because they claim that the banks cannot afford y2k remediation or have had problems.

(3) Companies stockpile a large inventory to cover for the first quarter of 00 and temporarily do away with JIT. The economy sinks in January from the lack of business.

(4) Scared depositors--individuals and businesses alike--start to yank out their savings and ask for cash. Here and there, a few banks start to run out of cash amidst hyped publicity. Suddenly everyone makes a run for the bank. The Feds have to declare a bank holiday until people are calmed down, nearly shutting down corporate America for a month.

(5) Less-scared investors make a run from the stock market and buy bonds or t-bills instead. This flurry of selling sends the dow into an automatic, computerized tailspin down to 5000. Because people have less savings, they spend less on discretionary goods. Another recession.

(6) Investors divest from emerging foreign markets, penny stocks, high tech start-ups, and biotech because they perceive these guys to be "too risky." Suddenly there are thousands of bankrupcies and loan defaults, which in turn collapse the hedge funds, creating a cascade of dominoes through the financial markets.

etc. etc.

I am not including the people who believe that they are on a mission from God to bring on the apocalypse, or those terrorists or foreign enemies who think that the BEST time to strike is around new year's 00.

So things are going to be really interesting from sheerly a social standpoint. That is why I think a scenario of less than 4 is out of the question at this point. Given the real likelihood of some bad technical failures that cascade from random and unforseen places, you (conservatively) can add on another 2-3 points to the y2k richter scale.

-- coprolith (coprolith@rocketship.com), September 08, 1999.


Moderation questions? read the FAQ