I have elected, effective today to fight the pollyidiots with a similar if not standardized question that they REFUSE to answer:
Name any software project ever completed on schedule, under budget, and bug free in the history of large scale software projects.
When this question is answered, which it never will be, I will fade quietly into the darkness; although I'll probably do that anyways since I have been warned by a friend at FPL. -- John Galt (jgaltfla@hotmail.com), July 28, 1999.

-- John Galt (jgaltfla@hotmail.com), July 28, 1999

Answers

That's too easy. none, but they'll agree that it's none.

they'll just say the failures that will occur will be bitr variety.

kinda like "i need money to go the grocery store" ---"sorry, our systems are down"--- minor "inconveniences".........

Sup-

-- SuperLurker (Slfsl@yahoo.com), July 28, 1999.

To the optimists, the enemy to be fought is date handling bugs in computer software. To the doomies, the enemy is *anyone who disagrees* with them.

JohnGalt, if you could get away with shooting every polly in the world, do you suppose the computers would run so much better? If not, why bother fighting them? Why not fix some bugs instead?

-- Flint (flintc@mindspring.com), July 28, 1999.

John,

There probably are no such projects. This brings out another important question:

Since we already have a whole world full of late, overly expensive, bug-ridden systems, why should we fear one more bug, or why fear this bug above others? (Okay, that's two questions. Sue me.)

Now, volume certainly is an issue. However, the volume of bugs in software has always been quite high, so why get our collective knickers in a twist about this bug and all it's manefestations? Is it really that much worse than the bugs that have already been kicking around in systems since the breed was invented? In some systems, probably, but not in all. I do not believe that Y2K is more or less serious than any of the other bugs running around the vast majority of systems. The only significant differences between Y2K and those other bugs are 1) it gets more press, 2) it is, as a specific type of bug, more prevelant than any other class of bugs such as an off by one error, 3) it just happens to be able to be tied to an event that everyone sees as somehow significant, namely the (percieved) millinieum transition and 4) it is generally easier to fix any given instance of the bug than other, less well-defined problems.

Is the bruhaha over Y2K much ado about nothing? No, of course not. A wide range of systems will have failures because of it. But guess what: A wide range of systems experience failures every single day as it is, and the world perks along just fine. If we examine it as a probelm to be dealt with it becomes no different from any other systems issue except in scale, and since the volume of people working on the problem is so much larger than on other problems I beleive we can expect the results to be approximately the same as we are used to seeing: Bugs, failures and reactionary patches that may not solve the problem entirely but keep things moving along until either a better solution can be found or some manager decides that the patch is sufficient and re-allocates resources to the next problem. Is it perfect? No, but then we've never had perfect before so why expect it now?

-- Paul Neuhardt (neuhardt@ultranet.com), July 28, 1999.

Thank you for the "polly" feedback.
However, I raise another issue.
When in the history, and this is my term (just for grins), of "bugdom" have all of the bugs converged on one particular SERIES of dates?
None in my programming experience.
Have a nice day. And most of all don't beg. It's shameful. Just rot anywhere but in my front yard.

-- John Galt (jgaltfla@hotmail.com), July 28, 1999.

Paul Neuhardt's back. Eeeeek!, they're all coming out of the woodwork. Wonder what's up.

-- number six (Iam_not_a_number@hotmail.com), July 28, 1999.

John, thanks for bringing these folks out of their crevices. It's always interesting to watch them squirm.

Ray

-- Ray (ray@totacc.com), July 28, 1999.

John, you said:

"Thank you for the "polly" feedback. However, I raise another issue. When in the history, and this is my term (just for grins), of "bugdom" have all of the bugs converged on one particular SERIES of dates? None in my programming experience."

Never, but then again, having a deadline announce itself at all is a novelty in the history of systems debugging. Generally, bugs appear on an uneven and often unpredicatable schedule. Sure, there are always more immediately after an implementation or upgrade, but from that point forward they appear seemingly at random and without prior notice. For a debugger, having advance warning of this sort makes Y2K a much easier bug to solve than most others. The convergence of the occurance of Y2K problems around a date (or a small cluster of dates) is an advantage from the point of debugging, not a disadvantage.

The main problem lies in how late some folks got started working on it.

"Have a nice day. And most of all don't beg. It's shameful. Just rot anywhere but in my front yard."

Not planning on begging (or rotting, for that matter) at all, and that's really the point, isn't it?

-- Paul Neuhardt (neuhardt@ultranet.com), July 28, 1999.

Thanks for the brief but useful post, John. BTW: your handle has a bit of hubris to it, assuming you are a fellow St. of O.

www.y2ksafeminnesota.com

-- MinnesotaSmith (y2ksafeminnesota@hotmail.com), July 28, 1999.

"Paul Neuhardt's back. Eeeeek!, they're all coming out of the woodwork. Wonder what's up."

I never left. It's just that it's been so long between threads that were worth replying to that it just seems I've been gone.

-- Paul Neuhardt (neuhardt@ultranet.com), July 28, 1999.

I agree Paul,

My whole career (14 years) has been one long series of bug fixes. Most bugs I look at are far more difficult than anything regarding a date.

With systems programs, my own systems (before release) have taken out whole mini computers.. oops. Hell, I've even taken out a Tandem fault tolerant mini-computer with bad kernal code (perverse sense of pride).

I've gotten calls from guys in Taiwan who relate to me in broken english some catastrophe that I'm supposed to magically fix.

I've pulled all nighters.. Who hasn't.

I still put up preparations, just because my power goes out even without Y2K, and we are starting to get earthquakes (new for the NorthWest).

Your right, software is not reliable for about a hundred reasons, that's why we are so good at fixing it.

-- Bryce (bryce@seanet.com), July 28, 1999.

I'm becoming a 'Paul' cheerleader!

Right again, Paul. Most bugs are NEVER pre-announced.

They happen:

- The day after the programmer has quit or left on vacation.

- The night before the 400 million dollar release.

- At the tradeshow in front of all of your potential customers.

- In the code that no one has touched for ten years. Long after anyone knows how it worked in the first place. Y2K is great! It's the ultimate scape goat. It's not my code, it's that date rollover in the OS Microsoft forgot to fix. By the way, here's a patch.

-- Bryce (bryce@seanet.com), July 28, 1999.

Sorry John,

Not only is there a *large* software (AND hardware) project that was done on time, it was done by the US Government.

It was bug free enough to finish the stated goal with 6 months to spare.

The total software needed for the project was many many millions of lines of code on myriad different platforms from mainframes down to the smallest computer in the world (at that time).

I don't know about the budget, though....

The project in question is NASA's manned space program.

Jolly is NOT a pollyanna.

-- Jollyprez (jolly@prez.com), July 28, 1999.

Jollyprez, your answer fits with mine (go figure that, huh!). Kennedy's "end of the decade" challenge was met using systems that were at the time (and still are) considered paragons of reliability.

And yet, there were a myriad of failures on each and every flight. Apollo 1 killed it's crew without ever leaving the ground. Apollo 7, a "text-book" flight, recorded hundreds of failures despite being labled the most sucessful space flight in history at the time. Apollo 11 astronauts jury-rigged switches in their lander with pens. We know what happened to Apollo 13. Apollo 14's Lunar Module system decided that it had received an abort descent signal even though the abort switch had not been pressed and the descent had not yet begun. Programmers on the ground had only one or two lunar orbits to find a way around the problem. Their answer: change the programming to make the abort system believe it had already fired so it would not automatically abort the landing when the descent engine fired. It worked, and Al Shepard got to be the only one of the Orginal 7 to make it to the moon.

Remember Skylab, the one-winged station that fell from the sky years before it was supposed to?

I haven't seen many statistics on shuttle systems relaibility other than to know that they are considered some of the most reliable ever created and still experience numerous recorded failures each and every time they are used.

My point is now and always has been that the assumption that Y2K will bring otherwise reliable systems down is based on the false assumption that the systems were ever reliable in the first place. And if we have always been able to do pretty well with buggy systems that we had to fix on failure with little or no warning, we should be able to deal with bugs that can actually be planned for.

The question in Y2K never has been "can we fix this?" The question in the past has really been "Are we going to fix it?" The question going forward is: "How well have we done to this point in fixing the problem and how well will we do going forward?" My belief is that the answer here will be pretty much the same as it always has been: "Not too well, but on the whole good enough to work with."

-- Paul Neuhardt (neuhardt@ultranet.com), July 28, 1999.

Its always fascinating reading through the old argument that goes something like, "OK, we won't have Y2K fixed in time, but so what -- lots of problems don't get fixed on time, and we get by just fine." As if, on a daily if not hourly basis, the systems that we CURRENTLY rely on are in constant need of fixing or they will not function or be reliable. That is ridiculous.

The truth is that, yes, there are always bugs, and generally they can be worked around. But they did not all suddenly occur at once, they were discovered at a reasonable frequency of occurence that allowed for workarounds to be developed. That is indeed software normalcy -- bugs occur at a reasonable rate, and there is a reasonable amount of resources and time to work the problem.

Y2K promises to deliver a tremendous number of bugs in a very small amount of time. Everywhere. With the integrity of our power, municipal water/sewage, banking system, telecommunications, etc., called into question.

"Y2K is just another software problem, we will take care of it when it happens" just does not cut it, folks.

-- King of Spain (madrid@aol.com), July 28, 1999.

King, you said:

"Y2K promises to deliver a tremendous number of bugs in a very small amount of time. Everywhere. With the integrity of our power, municipal water/sewage, banking system, telecommunications, etc., called into question."

No, it only promised that everywhere if left unattended everywhere. Now it only promises that where either still left unattended or only partially corrected.

-- Paul Neuhardt (neuhardt@ultranet.com), July 28, 1999.

Slow down, guys!

I'm still up there somewhere, wondering what made this thread so exceptional that Paul Neuhardt deigned to post....

-- Elbow Grease (LBO Grise@aol.com), July 28, 1999.

Pollys answer any and all questions. We have nothing to hide.

Regards,
Andy Ray

-- Andy Boy (andyboy666@hotmale.com), July 28, 1999.

And Paul Neuardt essentially is promoting what I have named the "Next To Last Pollyanna Defense": fix-on-failure. Which implies that all the money spent to date, all the time and resources spent to date, etc., was essentially unnecessary. That when Y2K failures occur, why our trained computer profesionals with rush right in and take care of things. No problem. LOL, LOL!

(You watch: As we move closer to the end of the year, there will be a pollyanna shift to "The Last Pollyanna Defense": that Y2K problems, though they will not be fixed nor tested in time, and though they cannot be fixed-on-failure, will nevertheless not be a big problems because ... they are fairly innocuous in nature. I.E., they never were that big a problem in the first place. [Or, as an alterative, the claim might be made -- at least in the good ol' U.S. of A. -- that they REALLY BIG Y2K problems got taken care of, and all that are left are the "small fry" problems that won't be of consequence. But this line of BS is a lot harder to swallow, so I'm predicting that the claim will be that ALL Y2K problems will be said to be trivial.])

-- King of Spain (madrid@aol.com), July 29, 1999.

Damn it. Who the hell left that stupid, annoying font type open. My post looks like one from that pathetic troll known as Andy Ray.

-- King of Spain (madrid@aol.com), July 29, 1999.

King,

I'm curious. How do you infer I feel there is "no problem" (your words) from my statement "Is the bruhaha over Y2K much ado about nothing? No, of course not. A wide range of systems will have failures because of it."

-- Paul Neuhardt (neuhardt@ultranet.com), July 29, 1999.

Oh come on, Paul N., re-read what I posted. Only the most dense of pollyannas still cling to the belief that Y2K is hoax, or that it will still ALL get fixed in time, etc., etc. Of course you believe that there will be "problems". Where we obviously differ is in:

1) The LEVEL OF SEVERITY of the problems.
' 2) The ability to apply FIX-ON-FAILURE to the problems.
3) The issue of whether problems that occur across all industries simultaneously will, in itself, impede fixing the problems. (I.E., it would take longer than if each problem were able to be fixed with everything else working normally.

Its practically August 1999, dude. Wake up and smell the coffee.

-- King of Spain (madrid@aol.com), July 29, 1999.

King:

Pardon my confusion, but in reading what you posted I found the following (emphasis mine):

"And Paul Neuardt essentially is promoting what I have named the "Next To Last Pollyanna Defense": fix-on-failure. Which implies that all the money spent to date, all the time and resources spent to date, etc., was essentially unnecessary. That when Y2K failures occur, why our trained computer profesionals with rush right in and take care of things. No problem. LOL, LOL!"

Since you used the words "no problem" in your representation of my position I was forced to assume that you meant what you said. So sorry to have misunderstood. I'll try to read your mind more clearly next time.

Another point I'm curious about. Exactly where did I here (or anywhere else for that matter) state, or even imply, that "all the money spent to date, all the time and resources spent to date, etc., was essentially unnecessary?" If you think about what I've said instead of just reacting to my presence here, you would see that I espouse quite the opposite position. The only reason I don't feel Y2K is likely to be a catastrophic situation is that I believe all the money, time and other resources spent on the problem to date were not only necessary but have been largely effective in dealing with the majority of issues.

Just as with any other systems bug, Y2K could have had serious implications if left unchecked. However, it has been attacked and dealt with in the same general manner that other systems bugs have been dealt with and I expect to see simlar results arising from that work. The only difference between Y2K remediation work and normal systems debugging work is that everyone has been working on many iterations of one bug instead of focusing on a myriad of different bugs as is normally the case.

I agree that you have highlighted most of the main differences in our viewpoints. However, I would add one more, and that is I never thought that most systems worked all that well to begin with and yet we all seemd to get along fairly well. You seem to hold the existing quality of systems in higher regard than I do, leading you to expect a greater impact from Y2K failures that do occur.

-- Paul Neuhardt (neuhardt@ultranet.com), July 30, 1999.