Ironically, the one reason I can find to be optimistic about Y2K is the widespread deceit and/or hidden-ness about the overall reporting of it. This gives me a faint hope that we are much further along in remediation than I expect and is why I personally still give a BITR a 5% possibility (see a's thread).

Against this lies the simplicity of understanding Y2K from an IT historical point-of-view (remember, I'm speaking broadly about global efforts, not about the efforts of various, specific entities):

1. Y2K budgets have steadily escalated (bad project sign).

2. Deadlines have shifted backward (bad project sign).

3. Testing is being deferred, reduced or scrapped for FOF (bad project sign).

4. Optimism when 75-90% code complete has been reached is taking hold (bad project sign).

These bad signs historically lead to late projects. They are the one SURE statistical metric within our industry! Alas.

Against these facts, it doesn't matter whether every media outlet in the world SINCERELY publishes "good news" reports of "progress" (progress is always being made in software). It doesn't matter how huge the disconnect is between public expectations of a "bump" and what will "really happen". The public can expect whatever it likes, it doesn't affect what will happen. Likewise, the markets. Likewise, CEOs (IT pros know how little they understand about their own systems).

All that matters is understanding how it could be (could it?) that these historical predictors won't come to pass this time.

The only argument I have ever read that is partly credible is the very one which ensures that Y2K impacts are going to amaze the world: the date is inflexible so everyone must and will finish.

While this represents a strong wish, software history also indicates that it is literally impossible to finish projects behind schedule by throwing more resources at them and, generally, it makes projects even later than they would have been.

As an intuitive rule of thumb, IT pros know this is why stated completion dates of Nov. and Dec, 1999 cause us dread. Except in the rare instances where the completion date is really September (and the geeks simply "told" mgmt it would be December), this really means that the GEEKS have said December .... and that means trouble.

The inflexible date on top of the predictive markers that are in place as of July 15, 1999 is extremely ominous.

Adding

5. The unpredictable impact of embedded systems

... into the stew is that much more ominous. Chillingly so, since there are bound to be strange impacts, even if the percentage of bad systems is as low as hoped.

Now, if it is all so simple, why has the IT press not picked up on this and made a huge stink?

a) Outsiders cannot imagine how boring maintenance programming is, in general, and Y2K is, in specific, to the tech press. Yesterday's systems, yesterday's languages, yesterday's news.

b) Related to a), just as very few computing entities have ever applied the lessons of the historical programming markers I cited above, likewise, IT journalists haven't either. Saying "why" would take a book and many books have been written. Outsiders simply have to take on trust that something about this profession profoundly resists the simplest possible measurements of "real" productivity and project progress.

c) The IT tech press, even more than the mainstream, ordinary press, is a creature of products and services ("advertisers"). Investigative reporting is mainly non-existent. Not minimal. Nonexistent. Everything is future-oriented (reviewing products and the "next cool thing"). Also, without putting too fine a point on it, and there are exceptions, most IT journalists are not themselves very technical.

To repeat and also conclude, this is the most important thing:

"Good news" at this stage is not trustworthy when coupled to the facts I state at top. These facts, admitting numerous specific exceptions, cannot be denied (if you think they can be, give it a shot). No doubt, some of the good news is authentic (SSA, having spent 10 years on this, can be taken as good news; some others as well). I only state that the news, broadly, is not trustworthy, since we are in the "optimistic phase" (itself a BAD, not a GOOD sign). I hardly need to point out that even the good news is usually accompanied by bad news anyway.

Y2K (as Cory never tires of pointing out) is not a poll-driven or "feel good" process. The projects are either being completed successfully or they are not. IT history and metrics says THEY ARE NOT. That is a deduction, but a very safe one. Unfortunately.

We can have good news right through December and still have Infomagic. Or a bump, yes, it is possible.

Contra the usual assumptions, I would be delighted to be shown the error of my ways. But it has to be by serious IT software professionals, not by the usual bozos.

Your serve, friends .....

-- BigDog (BigDog@duffer.com), July 15, 1999

Answers

Thanks, Bigdog. Every newcomer who reads this will understand a little more order from chaos.

My question to you is do you think the latest Y2k Federal Legislation concerning immunity from lawsuits has impacted the truthfulness of y2k reports and disclosures?

-- Jim the Window Washer (Rational@man.com), July 15, 1999.

Say BigDog,

if you get an answer, send it in along with your piece above. It would make a great WRP article.

You're probably tuckered out after writing all that, lying there on your side, still got some Hartz Polly-Off in your fur. It's pleasantly scented and safe when used as directed.

-- cory (kiyoinc@ibm.XOUT.net), July 15, 1999.

lol. Hartz Polly-Off. Need to sprinkle some of that on my screen.

-- Mike Lang (webflier@erols.com), July 15, 1999.

Jim - "the latest Y2k Federal Legislation concerning immunity from lawsuits" was perhaps going to be formally signed by Congress today and then sent to the White House, but (at least as of a few hours ago) it wasn't law yet. Stay tuned!

-- Brooks (brooksbie@hotmail.com), July 15, 1999.

Good boy, BigDog. Your logic is sound. Here, have a Snausage.

All I can add is an elaboration of your "feel good" comment. The US is probably in its most feel good time in its history. Thought, the tool which has actually our civilization to its current level, has now firmly taken a back seat to feeling in everyday usage. BJ Clinton "feels" our pain (among other things). He does not "know" our pain, he does not "understand" our pain...he "feels" it. (Which, incidentally, is about as bold-faced a lie as any other he has granted us, but I digress). The media, the feds, and the bulk of the well-conditioned population doesn't "think" about Y2k, or most other things for that matter, and they certainly do not "understand" the technical difficulties of Y2k or the potential ramifications Y2k may unleash. They've just been led to "feel" there will be no problems. Feeling is easier than thinking, more satisfying, less upsetting. Now, in 1999 America, doesn't that "feel" just about right?

-- Nathan (nospam@all.com), July 15, 1999.

Cory -- In all sincerity, let's hope this thread will provide the polly response to all this (and we doomers will agree to keep the flaming down), definitively. I'll cook it down into a doomer-polly broth, send the soup to you and you can pub or not as you'd like.

-- BigDog (BigDog@duffer.com), July 15, 1999.

BigDog, Puddintame here (a usual bozo.) Good post as always. I'm in Detroit on some ancient computer not my own. I can't even get my name to enter in this reply.

One of the things that really disheartens me is the media's grasping at straws. If there was some real blockbuster good news out there, it would be reported somehow. Instead we get a faint whiff of testing that has not completely exploded in the testers' faces and it's reported as bolckbuster good news, eg. the defense department procurement systems testing. (An apparent exception to this is the NYSE trading systems testing. Although I've seen no details, it was reported as extensive system wide testing.)

I hope to be able to catch up on all posts when I return home. Of course, to do that I'll have to leave this compliancy mecca of Detroit. What a joke that is. If this place is ready it's only in the same sense that Russia is ready. Everything is already fubar so what else can go wrong.

-- (achillesg@hotmail.com), July 15, 1999.

Damn. All too much truth here.

-- Lane Core Jr. (elcore@sgi.net), July 15, 1999.

Haven't we been over this enough yet? Surely by now it can be taken as a given that *nobody* will be completed on time. The best will be in various stages of 'close', the the worst will have to cross their fingers and take their chances.

Net result: a global mess. Localized but widespread disasters of varying severity. Lots of inefficiency, lots of fire drills, lots of delays and screwups and downtime. The Age of Exasperation. Manageability level unknown, duration unguessable. YOU as an individual may hardly be affected, and you may be devastated. The Big Picture will work itself out one way or another. Prepare for the Little Picture, since you'll have to live through it.

-- Flint (flintc@mindspring.com), July 15, 1999.

Flint -- Even on this forum, the primary argument seems to be over that very question. Koskinen et al as much as insist with the "three-day prep" argument that almost everything WILL be complete. That is the entire tenor of the media coverage.

Granted, anyone reading the articles carefully could well come to a different conclusion, but one has only to consider what a "casual" assent to my post above would mean for global preparation to see the huge gap between "me" and "them". Yes, I believe the argument I make IS quite simple. But it doesn't seem to be "simple" to most people?

An implicit corollary to my post is that problems will be percolating for months, at least, after rollover across all IT domains around the world. One could argue that this means "recession" instead of "depression" and, indeed, we'll find out. But, again, this is not viewed as a "trivially correct conclusion" by the media or by many on this forum.

Or am I missing something?

-- BigDog (BigDog@duffer.com), July 15, 1999.

Big Dog:

I think you're asking two different questions here, so I'll try to answer both:

1) The *extent* of the problems, as I described them, is purely my own subjective impression of what we face. Clearly, assessing this extent is the focus of this forum, and while I take it for granted that my expectations are correct [grin], I can certainly understand how those not so prescient as I may (quite mistakenly, of course) feel differently.

2) My take is that it is indeed considered trivially obvious in the outside world that there will be problems, at least by those who have considered it at all. Koskinen's warnings, to the vast majority, seem excessively serious. Sure, something will probably go wrong somewhere (the feedback I get says), but it won't affect me any more than I'm affect by life's typical imperfections.

And as you've said many times, the Bad Signs you see don't necessarily preclude a Koskinen level of public impact. They would appear to guarantee Holy Hell within the glass rooms everywhere, but most people have never even seen such a room. How much these difficulties will escape these rooms is still highly problematical.

And I might mention (as a somewhat off-topic footnote) the latest edition of EE Times is devoted to predictions for the semiconductor industry for the next year. There are articles by 19 CEOs of major players in this industry, and NOT ONE mentions y2k at all. The gist is that the industry's been in the doldrums for a year, but is coming out (3 straight rising quarters) and will shine and soar for the next 12-36 months, no doubt about it.

At the end is a long editorial summary by the EE Times editors, and they mention y2k in ONE sentence. This sentence says (not an exact quote, don't have it with me) that the net impact of y2k being as likely to be negative as positive, it's considered neutral and therefore factored out as an active consideration of future trends!

So maybe you're right. I may sound generally optimistic in the context of this forum, but I'm a wacko in real life.

-- Flint (flintc@mindspring.com), July 15, 1999.

Interesting post, BigDog. Much of it certainly fits what I gleaned from reading a few of Capers Jones's books six months ago.

Of course, a key question re all those companies and agencies that don't get done in time is, "How close is good enough to still muddle through?" It seems very hard to get any sort of handle on that issue; I sometimes wonder if experienced IT people inside each particular company really know the answer. Also, I sure wish we could get more reports and info on all the supposedly "noncritical" systems out there. It seems as though almost nobody is talking about those systems any more, and consequently I have no idea how much is getting done on those systems in both the public and private sectors. Indeed, the overall quality of Y2K information these days seems extremely shoddy despite the glaring importance of the subject.

I've no idea how all this is going to turn out, but I'll offer a couple of cultural and psychological reflections, for whatever they are worth. You probably remember C. P. Snow's book, originally published some forty years or more ago, about the "two cultures" in academe: the scientific culture and the liberal arts culture. Snow lamented the fact that the two cultures didn't communicate much with each other; well, they still don't. (OK, at places like St. John's they might.) When I used to teach at the U. of Illinois, I sometimes tried to bridge the two cultures because, although my primary training has been in the liberal arts (doctorate in English, specifically Amer. Lit.), I've also had various strong scientific interests (esp. astronomy and physics) since early adolescence--plus other interests, like economics, which are rather in that messy "no man's land" between the two cultures. But my experience at Illinois was that the Great Divide between the two cultures was as great as ever.

Politicians, lawyers (and most politicans were trained as lawyers), journalists, and many other social and govt. professionals are trained to regard the word, the perception, the image, the idea, as paramount. A lawyer wins a case by persuading a judge or jury to see things his way; yes, "evidence" matters a lot, obviously, but so does rhetorical skill (something else that I used to teach at Illinois). Politicians and journalists tend to think in terms of social constructs, which are also largely based upon words and perceptions.

In many of the high-powered academic liberal arts circles that I used to inhabit occasionally, one finds great emphasis upon modern theoretical interpretive trends like Derridean deconstruction and various other post-structuralist theories. Most people, including politicians, lawyers, and journalists, who are outside of these academic circles don't really understand just how these theories work (their foundations are incredibly abstract and arcane), but there has been a vague dissemination of the general "sense" of these theories into the public arena over the past two decades, and especially in the Nineties. What gets promulgated, in a rather loose and often subconscious way, are such notions as these: "truth" is largely relative and cultural; virtually anything can be deconstructed (re-interpreted) to fit one's own perspective, idea, or agenda; society itself is but a fluid mental construct; etc. Mr. Clinton's rather amateurish attempt to re-define the meaning of the word "is" is simply a small part of such theoretical, linguistic, and semantic trends--a somewhat grotesque and comic microcosm of what has been gradually transpiring in American culture as a whole. In economics, we have seen the re-interpetation of stock valuation models to deconstruct traditional economics and justify a sky-high stock market.

Many of these theories originated in France in the 1950s and 1960s, largely as a reaction by thinkers like Derrida, Lacan, Foucault, etc., against authoritarian teaching methods and conceptualizations in that country; but since the early Eighties they have come into their own in America. Partly, I suspect, the growth of these theories in America came as an academic counter-reaction to the "Reagan Revolution." But also American society, which has long emphasized images (advertising, television, etc.) and political "spin cycles" (just at lower speeds in the past!), was clearly destined to be receptive to such theories once they were watered down for popular consumption.

And frankly, such theories aren't always and inherently "wrong" or "useless"; they can lead us to question cultural and social dominance by particular groups, for instance, and to re-evaluate whether many things we've considered as "absolutes" (or at least "standards") in the past really are such. In my own field, such theories have led to some valuable re-analysis of the literary "canon," for example. The problem, of course, is that we also live in a physical world of facts. You can deconstruct/re-interpret "gravity" all you want, but if you fall off a cliff you are going down. It is often easy to forget or minimize the physical facts and constraints of existence. The more complex a society, the more complex and pervasive the comforting illusions, the ideational structure, constructed for the benefit of its members--and when you factor in the trends discussed above, you understand why it is often very difficult nowadays to confront directly unpleasant facts. Y2K represents a "physical fact" in the sense that you can't deconstruct/re-interpret a computer system that doesn't work so as to make it work. The closest you can come--and at least some govt. agencies and companies have evidently done this--is to re-classify (deconstruct) "mission-critical" systems as "noncritical" ones.

As I said before, I don't know just how Y2K is going to turn out. But the above paragraphs might explain some of our peculiar cultural and social obstacles in confronting the subject directly and openly.

-- Don Florence (dflorence@zianet.com), July 16, 1999.

Whoa, Don.

Your statement takes me back to my worn copy of "The Western Canon" and Harold Bloom's statement "Hamlet did not have an Oedipus complex, but Freud certainly had a Hamlet complex, and perhaps psychoanalysis is a Shakespeare complex!" (Grin.)

Getting back to Y2K. I too don't have any idea how bad things will be next year. But if I were a rationalist living back in the 16th century and someone had argued that a band of few hundred adventure seekers with a few muskets and horses could take down a major civilization with tens of thousands of soldiers, I would have said that the idea of such a conquest is idiotic. Of course I would have been wrong. (Check out a good history at your local library on Cortes and the conquest of Mexico.) And the last Tsar thought that the idea of revolution in Russia was idiotic, too.

-- Alexi (Alexi@not-in-the-dark.com), July 16, 1999.

Hey, Cory --- Apparently, we're making progress. No one disputes the reasonableness of my analysis: Y2K remediation, taken as a whole worldwide, will fail (again, that is obviously a deduction/prediction at this point, based on my analysis). As recently as six months ago, many declared it would succeed.

BTW, I have no problem with folks saying it will fail but the global impact will be minor (actually, I think they're nuts, but I'm saying one could debate that separately), you know, a three-day thingie.

I'm surprised. I guess everyone has become a doomer, TECHNICALLY speaking. Maybe we're finally beginning to agree on the basics. No sarcasm intended, that would be a big step forward, I suppose, though no help to stopping-the-train-and-getting-off.

-- BigDog (BigDog@duffer.com), July 16, 1999.

Big Dog

What a GREAT post! I've been reading this stuff for months and I was beginning to get info overload. Your post laid out exactly what my concerns have been. I am not a geek but have been surrounded by them my whole life. I know how software projects work - or rather why they often don't.

Could you be a little more clear about the following statement?

"4. Optimism when 75-90% code complete has been reached is taking hold (bad project sign)."

It is a little confusing in the wording. I understand it to mean that it is a bad sign when people are optimistic about reaching the target date when only 75-90% of the code has been completed. Is there some sort of delusional phase in programming? If a project was going well and only 75% of the code had been completed wouldn't people be optomistic then too?

-- R (riversoma@aol.com), July 16, 1999.

Good programmers are never optimistic during a project because they know that the only "good product" is a "shipped product" (ie, TS can HTF anywhere along the way and crash schedules).

The majority of programmers, like the majority of people (moi included) are more-or-less sloppy and mainly interested in getting the cool stuff "mostly working." 75-90% code complete.

Regrettably for schedules, the last stuff is always the hardest and the least fun to finish (we're not even talking testing, because that follows code complete). The reason, drastically oversimplifying, is that users cannot/will not put up with rough edges that programmers handle during code time intuitively because they "know how it all works". Most programmers sincerely can't understand why those rough edges bother users at all!

Closing off all the application behavioral "gotchas" so that working stiffs can get their work done on time without quitting their jobs in frustration or hurling the terminal/computer into the trash usually takes as long as the first 75% or longer.

Of course (this is a trade secret), the hackers never really HAD completed the first 75-90%, they just thought they had. And because management is all too willing to report, wishfully, that projects are almost done (helps with their bosses, securing more budget, etc), everybody plays the game.

Doesn't matter CRITICALLY when the dates can be slipped.

When they can't? Gee, we'll soon find out, won't we?

Come on, geeks, PLEASE tell me I'm blowing smoke. We're getting down and dirty here with life the way it really is in the trenches. Yes? No?

-- BigDog (BigDog@duffer.com), July 16, 1999.

Don -- Answer to your interesting post coming up later today.

-- BigDog (BigDog@duffer.com), July 16, 1999.

Flint, say it ain't so!

The Age of Exasperation

You mean we aren't already in The Age of Exasperation?

:-)

-- Lane Core Jr. (elcore@sgi.net), July 16, 1999.

Big Dog,

What you are saying makes total sense and is very enlightening. I am wondering how it is that my father, step-mother and step-brother - geeks all - can be oblivious to this kind of logic. Why is it that many geeks who have been writing code for a decade or more don't think Y2k is any big deal. What are they thinking? Is the project just too big for them to grasp intellectually? Are they so dazzled by the dollars being spent that they think all the money in the world will suddenly make testing the shortest part of the process? Even though they KNOW all about testing? Even though at that moment they are walking around muttering because their own tests for their own projects have revealed so many bugs that they don't know where to start fixing?

I truly don't understand the Polly Geeks. I understand why Joe and Barbi McSixpack don't get it. Writing code is a pretty arcane process to non-geeks. But how can any geek look at me with a straight face and say they will fix on failure? With what? With who? How the heck are they even going to find and test all the damn bugs if you CAN'T RUN THE FRIGGING PROGRAM BECAUSE OF THE BUGS?????????

Am I missing something here? How many crashed interactive programs, how much buggy hardware does it take before it is impossible even to figure out which program is buggy? Let alone what is actually wrong with that program/system?

-- R (riversoma@aol.com), July 16, 1999.

Riversoma:

After only a little thought, you should realize that you are describing a fairly common problem. Almost any program you run after your very first compile of shiny new code will hang horribly -- for a hundred different reasons (depending on program size, of course).

At my level, the problems are found with oscilloscopes, logic analyzers, in-curcuit emulators, etc. At a higher level, the code can be run under the control of a debug program. There are debuggers that can view the operation of the program from the viewpoint of the CPU, or any number of buses, or the network. There are time-domain reflectometers to find network cabling problems. There are 'core dumps' and programs to analyze these dumps. One of my key job duties is to write diagnostic software (often one-shots) to find specific problems. There are 'data sniffers'. Indeed, there are thick catalogs of debugging equipment and programs at our disposal.

Without all these tools, we'd have been lucky to get the first "hello world" program running -- and even if it did run, we wouldn't know if everything went right. So this isn't the problem you think it is.

-- Flint (flintc@mindspring.com), July 16, 1999.

Flint,

I do understand that point and upon re-reading my post I can see why it came across that simplisticaly. The point is this. If you have ONE buggy program then you have the luxury of being able to use those diagnostic tools. However what if you have 4 interdependant seperately written and designed programs which must all run perfectly in order to create a final product. What if you don't know which of these programs the bug is even in?

What comes to mind is a job I had at a radio station. There were 4 separate programs written by 4 seperate software companies. All the programs had to work for the station to run. There was one to schedual spots and do invoicing, another to schedual music, another to play music, one to play spots. When live DJs were on then we would do a manual override on the music schedualing computer. Late at night and on the weekends the station would literally run itself.

I cannot imagine what they are going to do in Y2k. These systems are all very date sensitive all the programs must work together and there are no industry standards. Each software package we used was one of dozens there were to choose from. There is no Bill Gates of the media software industry to get everyone to conform to some degree.

When the radio station shuts down as I fully expect it to (last I heard they were making lots o jokes 'bout that silly Y2k scare) then they will be toast. They will have no idea which of their 4 systems is in trouble. So lets say that they manage to get the problem software to its creators. Lets say they even CAN fix it. Lets say (miracle of miracles) all the software comes back after FoF. Its all hunky dory. All the microchips in the whole station happen to be compliant and there are no power outages up here in the rural northwest. How will they ever get the programs to be compatible again? How will they get Humpty Dumpty patched up and back on the dang wall?

Well they can't. So they decide to get new software. It is installed and staff stays up all night training. How long has all this taken? Weeks? Months? How does a radio station make money if is not on the air? They will not be able to play any spots or music even with a manual override if the system has crashed. They will have signed many prepaid contracts for the time after 1/1/2000. They will have to refund all that money.

I happen to know that this radio station totters always on the brink of bankruptcy. I had my own bookkeeping business for many years. Most little businesses run close to or in the red ALL the time thats why they are STILL little!!!!! They have no money to expand!

This is just one business in one industry. It is a fairly small company with less than 10 employees. They will never make it. There are thousands of little businesses like this. Limping along from purchase orders to invoices. Always barely covering payroll. Sweating out the quarterlies to the IRS. They will NOT make it.

Well so what? So a lot of little fish don't make it? You can't have big fish without little fish. The little businesses produce things the big businesses need. If it is only the income which the big businesses squeeze out of a little business employees paycheck. Many of these little businesses are software companies. Other little businesses are dependant on them - as well as big ones.

Again what comes to mind is the spot schedualing software. It was produced by a tiny little company but is used in some major markets. They are NOT Y2k compliant, ready or able. They are toast and so is everyone they do business with.

Diagnostic tools are great when you are using them in-house and have some ability to isolate the software and change it without affecting anything else. Perhaps radio is more interdependant and convulted than most businesses. Maybe power plants are much simpler to run than a micro-market radio station.

-- R (riversoma@aol.com), July 16, 1999.

I'm not sure what Flint was getting at but there is a qualitative difference between module-level debug and the maintenance of > 1M line software applications.

Very few programmers appreciate the step-up in complexity that is entailed by enterprise-size systems. Our own 'a' is quite right that, even apart from Y2K, many systems are reaching levels of complexity that augur poorly for their survival. Cory's opinion has been that, without Y2K, we are still looking at the failure of not-a-few huge systems over the coming years (beyond the current percentage, naysayers) due to this phenomenon.

Moreover, very few programmers understand testing. Testers and programmers are temperamentally and intellectually distinct in their abilities and mindsets. Within a very large industry, these are two distinct professions and career paths.

To use an analogy, many people today "think" they understand computers and programming because they've cobbled together a small Visual Basic application. In a sense, they do. In another sense, they're clueless. Unfortunately, "everyone knows" has become the bane of Y2K remediation.

-- BigDog (BigDog@duffer.com), July 16, 1999.

Don -- The really pervasive ideas (for instance, Marxism in its day, Freudianism and, now, deconstructionism) affect most of our cultural life even though we are largely unaware of their impact. Maybe, especially because we are unaware.

There is no doubt that Clinton treats everything as a "language game" (cf Wittgenstein as well as Morris/Carville). Likewise, Koskinen with Y2K. And Greenspan with the markets.

In a sense, Y2K is a language game in the real sense of that word! Man-made languages with a grammatical construct (dates) that can be "gamed" (budgets, project management, press releases) around the notion, "is this construct broken?"

Unfortunately, as you point out about gravity, some language games are tied to external phenomena. In this case, Y2K is not just an electronic token that can be deconstructed but a way we conjure whether oil wells will work, Medicare will send checks and nuclear missiles will remain safely in their silos!

In a sense, Clinton-Koskinen (Greenspan too) are betting that Y2K can be gamed away from wrecking the "confidence" of the markets (the real game in town), that is, that the language effects will be far more severe than the "gravity effects".

They're wrong about that. And being wrong means that ordinary folks who have failed to prepare because of THEIR Derridean mindset will be hurt next year.

-- BigDog (BigDog@duffer.com), July 16, 1999.

"Is there some sort of delusional phase in programming? If a project was going well and only 75% of the code had been completed wouldn't people be optomistic then too?"

I know this sounds wacky, but I've found it to be true:

The last 10% of a project takes 90% of the time.

I'ld love to get into this, but I've got a big day tomorrow. A 15 year friend is getting married, and the limo is picking us up at 9:15. I gotta crash.

BD, and any other pros out there, please explain my remarks. Thanx, and have a great weekend!!!!! <:)=

-- Sysman (y2kboard@yahoo.com), July 16, 1999.

What a great thread.

Deep. Thought provoking. Reminds me a bit of this forum 7 months ago.

'Gotta tell you though, the one quote that stood out for me as "soundbite" quality was this one:

"How does a radio station make money if is not on the air?"

The answer IMHO, is that it never hints to ANYONE that it suspects might NOT be on the air. If true, that approach could speak volumes about every other business/government entity in the country and its current reporting status.

You guys are good.

:)

-- FM (vidprof@aol.com), July 17, 1999.

Big Dog:

I hadn't intended to respond to this thread at all because I don't think it serves any purpose to extrapolate on experiences to include the whole world, not to mention that I've never looked at Caper Jones' book, nor studied these metrics that you present at the top of the thread. However, having 30+ years' experience in large systems, I have to take exception to some of the generalizations you've made here:

"Good programmers are never optimistic during a project because they know that the product" is a "shipped product" (ie, TS can HTF anywhere along the way and crash schedules)."

This one really struck me. I remember quite distinctly two different projects wherein I looked at the work that needed to be done and the deadline for the work and said flat out, "Is this possible?" The interviewer at the client sites in both cases was the project leader. His response was simply, "We're GOING to do it." Do you really think Private Ryan would have been found if "good" troops weren't optimistic?

Of COURSE everything that CAN go wrong WILL go wrong. On the first of the two projects I mentioned, we had a deadline 6 months out that seemed insurmountable to START with. It was a maintenance project that included perhaps 400 programs of varying magnitude in languages that included Assembler, COBOL and Eztrieve. We had 6 people. The specs changed so many times that each day we asked "What's the change today?" We laughed until we cried because it just didn't seem possible or reasonable that these folks expected what they did, but we all worked hard and rolled with the punches. This project included perhaps 10 fields that needed to be expanded in perhaps 400 programs that used perhaps 20-30 files. Of course the spec changes were always on the field sizes. The merging companies just couldn't agree on final lengths. In addition, each company used various subsets of these fields to mean different things. Yes...VERY frustrating. To top off the spec changes, a racoon traversed a high-voltage wire that threw the computer center out of business for several days, and a hail storm collapsed the roof of the computer center taking it out of business for an entire week. We didn't even bother to test changes. Why would we test one day just to change it all again the next?

I don't think we had more than 2 weeks to test all those changes, modify all the JCL, etc. Did we do it? Yep. We spent 36 long hours working on that final test, and had the IT director standing before us in the morning waiting to verify the results. Following verification, we spent another 24 hours moving the programs to production and babysitting the production system.

Good programmers aren't optimistic? Good programmers LAUGH at adversity.

"The majority of programmers, like the majority of people (moi included) are more-or-less sloppy and mainly interested in getting the cool stuff "mostly working." 75-90% code complete. Regrettably for schedules, the last stuff is always the hardest and the least fun to finish (we're not even talking testing, because that follows code complete)."

Here again, WHO are these programmers? Are they burned out? Do they no longer enjoy programming? The hardest is the LEAST fun to finish? What exactly is the hardest? Do you mean the most complex? The teams I've worked with knock those out after gaining familiarity with the file structures and the system in general. The most complex typically DRIVE the data that enters the least complex. At least that's been MY experience.

"The reason, drastically oversimplifying, is that users cannot/will not put up with rough edges that programmers handle during code time intuitively because they "know how it all works". Most programmers sincerely can't understand why those rough edges bother users at all!"

Here again you're making generalizations. I recently worked on a project where the code and results were so flawed that I asked the manager, "Your users actually put up with this?" I've seen quite the opposite to be true. Have we worked on separate planets? [grin]

"Very few programmers appreciate the step-up in complexity that is entailed by enterprise-size systems."

How do you think so many of us have maintained careers on large systems if we couldn't appreciate the complexity entailed by enterprise-sized systems? "Moreover, very few programmers understand testing. Testers and programmers are temperamentally and intellectually distinct in their abilities and mindsets. Within a very large industry, these are two distinct professions and career paths."

We don't? They are? These generalizations are becoming ridiculous, Big Dog. It IS true that some industries have evolved in recent years to include specialized testers. I wouldn't have that job for the world. I've only worked at ONE client site where they came in. They documented and tested the system there AFTER we had completed OUR testing.

-- Anita (spoonera@msn.com), July 17, 1999.

Anita --- Of course, I'm making generalizations! I hate to tell you, but this is legitimate when one is discussing subjects "in the large", whether Y2K globally or IT as an industry. Even a 500-page book on these subjects would make many generalizations. The question is whether the generalizations are true.

Actually, the question on this thread is whether my original post is correct or can be opposed on equally credible grounds.

No question, there are programmers who fit the profile you described above. I have met them myself. Heck, I've been involved in projects like the one you describe. You wouldn't be on this board if you were ordinary, frankly, you're a kook. The doomers are kooks. The pollys are kooks. That isn't an insult. We're off the normal sections of the bell curve.

Now consider when "9 to 5ive ers" in the profession (the vast majority) are involved in scut Y2K projects that look-and-sound a lot like what you described, only they are underbudgeted, politicized and otherwise fitting the criteria that we see PUBLICIZED about Y2K worldwide? All as I described at the top of this post. Y2K as it is, "generally" speaking, "worldwide" speaking.

I'm hoping someone will give that a shot, but it can't be done. Not because it's "just my opinion", but because the one lesson IT as an industry has learned historically is that these are the predictive markers.

BTW, any place where programmers test their own code has their head up the wrong you-know-what and any manager who permits it is a fool. Does it happen? Yup. Millions of places. In the circles I have worked, dedicated testers have been the rule for twenty years. And they would no more associate with programmers than vice versa. THAT'S the point. That's the way it should be, if we want to end up with a profession that does quality work.

-- BigDog (BigDog@duffer.com), July 17, 1999.

FM

Truth - every and I do mean every - small and mid-sized company in the world is bluffing about their ability to deliver goods and services next year. Its like playing liar's dice.

For those of you who have never had the dubious distinction of hanging out in a bar and playing Liar's dice it works like this. Each person has 5 dice. Player 1 rolls and doesn't show player 2 what they have. Then they say "I have three 5s. Well maybe or maybe not do they have three 5s. Maybe they have only two 5s but they are betting that player 2 has at least one 5. They can say whatever they want because they are self-reporting. Then player two raises the ante. "I have four 2s" well maybe they do maybe they don't. Eventually one of the players calls it and the last person to play has to prove their statement. If they were telling the truth and if they acuratly guessed what dice the other player had that could be counted in their total - then they win that round and keep all their dice. If they were bluffing or were wrong about what the other guy really had then they lose a di. This continues until one player loses all their dice.

To bring this out of analogy world and into reality. I am a glassblower. I work on a torch and I use borosilicate glass (Pyrex) tubes and rods. I have one client who markets all my work for me (on e-bay!) and about 5 separate vendors I get raw goods from. Clear glass, color glass, propane, oxygen, tools. Not a single one of my vendors is y2k aware - let alone compliant.

Up here behind the Redwood curtain we have one supplier of commercial oxygen. They have a regional monopoly and get to charge folks like me twice as much for oxygen than I would pay if I lived in the SF bay area. Last month when I asked the owner (a down-home-good-ol-local- boy) what they were doing about their Y2k remediation he just blinked at me. "Y2 what?" was his response.

I will not be able to get oxygen in 2000. Nor will I be able to get propane or Pyrex. This is why I am -uh- diversifying my interests. Honing up on other skills which I suspect will be VERY valuable. But I digress. The point is this. I will not be blowing glass in 2000 and neither will anyone else. There are thousands upon thousands of glassblowers making functional ware. (To those who are curious - Pyrex tubing is used to make scientific glassware as well as art) All of those glassblowers will be out of work. All their customers will be out of product and so on.

Do I mention any of this to MY customer? HAH!!!!! Gimme a frigging break. I'll just call him up and say "Hey I won't be producing any work after 1/1/2000 so don't invest anymore energy in my line of work. Stop promoting my glass 'cuz there won't be no e-bay soon anyway."

Certainly the world will continue to spin without my glass line. But I have to wonder what will be the economic effect of EVERY glass blower in the world unable to produce? Thats just ONE TINY industry!!! I could continue to cite real world reasons why every small business I have ever worked for or in won't make it in a Fix on Failure world. Once those businesses fail they will not be fixable. Ther is NO FoF for small to mid-sized companies. They will go down and they will take everything else with them.

Meanwhile we are all playing Liar's Dice.

-- R (riversoma@aol.com), July 17, 1999.