Is there proof the Y2K problem can be solved?greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread
Almost everyone assumes the Y2K problem can be solved. Why do we make this assumption?
In theory Y2K can be solved if given enough time -- I'd estimate about ten years of work, given the shortage of programmers and embedded system technicians.
In practice, soiftware bugs are never completely removed even when the sofware development projects are far smaller than Y2K.
Think about a large organization with over 100 million lines of code. Little Y2K remediation had been done before 1997. The organization shares data with over 1,000 suppliers/partners and also has over 10,000 embedded systems in buildings/factories/etc. Is there any evidence other than wishful thinking that this "system" consisting of the organization, all the suppliers it shares data with and all its embedded systems can be fixed in two years? Is there any evidence the "system" can be fixed in three, four or even five years?
I submit there is no evidence at all because not one large organization similar to my hypothetical example has completed their Y2K job. I must conclude our assumption that large organizations can fix Y2K are opinions based almost entirely on wishful thinking.
We still need data from several completed Y2K projects at large organizations to establish a pattern or trend for even a rough estimate on when Y2K work in the United States will be completed, if ever. I'm trying to be logical, not pessimistic, but without data, every statement based on the assumption Y2K can be fixed is nothing more than wishful thinking.
Please offer me some hope: I want to see data, numbers, facts -- not more words based on faith, hope, and wishful thinking. I get very concerned when I observe the Social Security Administration started fixing only 33 million lines of code in 1991 and is still not done -- not to mention the additional 33 million lines discovered in late 1996 and later declared to be non-mission-critical (short for "can't get it fixed before 1/1/2000 so we'll claim it's not important").
If there are no Y2K examples, how about telling me about ANY large software project (over 10 million lines of code) that was completed on time, delivering the originally planned functionality, and with 98% or more of the bugs eradicated before release? I'm waiting patiently and promise not to panic before 1999!
-- Richard Greene (email@example.com), June 23, 1998
Visa is the only large company I've heard of that has a pretty good handle on Y2K. Our local electrical utility has pretty much solved it's internal software issues, but their embedded systems issue is a bear. There will be many, many companies that solve this thing. The problem is the interconnectedness of everyone.
Some (not many, but some) small companies can revert to paper without much hassle. But, the vast majority of middle to large companies simply can't do that. Besides, no company can survive if the phones are down for any length of time. And phones and power are interconnected (see latest testimony from the telecommunications industry). This means that even if your bank is compliant, nothing is going to happen if the phones are down.
Anyway, this is old news. The key is interconnectedness. That's the tripwire into Y2K oblivion.
-- Pastor Chris (firstname.lastname@example.org), June 24, 1998.
This is to Pastor Chris....ummmmm....how can we revert to paper if the paper mills and all their suppliers are out of commission because of the power being down?
-- Sheila Ross (email@example.com), June 24, 1998.
When organizations used to be run by shuffling paper, they had rooms full of preprinted forms. They had manual typewriters. They had carbon paper. They had hand-cranked mimeograph machines. They had pencil sharpeners. They had hand-crank adding machines. They had green eye-shades. I don't think most organizations keep those items in supply much anymore. <<<<<<<<>>>>>>>>
-- Dan Hunt (firstname.lastname@example.org), June 24, 1998.
I guess I should add pencils, paper, manual typewriters, and hand cranked duplicating machines to my list of barter items.
Seriously, it does point out how bad things might get.
-- Rocky Knolls (email@example.com), June 24, 1998.
Richard, Don't give up hope just yet. There are quite a few of us who have successfully solved our internal Y2k problems and are well on our way to developing contingency plans for those things which are out of our control. You want facts, I want facts, all God's children want facts. However, the lawyers are laying in wait for the fools who are wreckless enough to provide blanket assurances and so all of us in leadership positions in Y2k programs are put in the position of guarding our words. You know about the legal impediment to full disclosure but do you have to live with it? Talk to people in your area working on the problem and you will find many who have the situation well in hand. Don't categorize me as a Pollyanna because of these remarks but I would like to lend my 30 years of IT experience in multiple industries to the voice of guarded optimism. Here are some facts worth considering before you build your cabin in the outback. First, some really big name Y2k experts like to keep harping on the delivery success of typical IT projects as an indicator for Y2k projects. This is a ridiculous comparison because the major delaying factor in development projects is "scope creep" and scope creep is not a factor for a programmer who is making a Payroll system Y2k compliant. In fact his specs are clear and cast in concrete whereas if he were developing a new payroll system then he would have to deal with the accounting department continual adding functional requirements and causing him to retrace much of his coding efforts. Anyone who does not know that scope creep at the functional level is the main reason for large software projects coming in late is not qualified to claim expertise in software development. So let's call this Myth #1: Y2k projects will be late because most software development projects are late. Now Myth #2: It has taken company A five years to modify ten million lines of code so it will take company B five years to modify its five million lines of code, ergo company B just beginning in 1998 will fail. Wrong, wrong, wrong, stupid, overly simplistic conclusion! Company B will have the following over company A. Company A (and all of us who began actually working on this problem before the experts began writing about it) did not have the benefit of the experience of similar companies which had solved it, Company B does. Company A could not benefit from the tools that have recently become available, Company B can. I can tell you from experience anyone who was working on this problem five years ago probably did not have much support from senior management and had to handle this as a "skunk works" project. Anyone starting now should have management's undivided attention and the enormous advantage that provides in expediting a project. I don't want to be too critical of folks who make their living talking about this problem but the above myths are becoming increasingly annoying to those of us who make our living solving problems. There was a need for hyperbole to get the message out and I assume that the professionals who have exagerated some of the issues did so with the purest of hearts. However, the word is out now and those same people will soon be called upon to help prevent the panic. So the tune on the rubber-chicken circuit will change soon. With Y2k hysteria going mainstream the big bucks on the speaking tour next year will be made by the folks who can debunk the myths and calm the masses. Richard, we won't solve all the problems but neither will our society melt down over this bug. It's not too late. Now that we have gotten the mule's attention with a two-by-four we need to attach him to the plow before the rains come. Have a great new millenneum, Don
-- Don Harlow (firstname.lastname@example.org), June 24, 1998.
Richard, I wrote my first response rather hurriedly and would like to add a few remarks. First, I understand that there has been a need to hyperbolize this problem and I apologize to anyone who would be offended by my comments regarding certain methods of sounding the alarm. I still think we are at the point now where awareness is no longer a big issue, at least not in the USA. Yes there are small businesses which still aren't doing enough and some will fail but the awareness train has plenty of momementum now and what I am seeing down here in the trenches is growing anxiety among the civilians which may be temporarily useful in forcing some leaders to act, but will ultimately cause a lot of uninformed folks unnecessary grief. It is not too late to fix most systems, even large systems. If we continue to propogate the myths of which I expressed concern then we will convince a majority of people that it is too late and then we will have created the environment to assure our failure. I have noticed that any attempts at voicing optimism are not usually well received on this forum but I do believe the time has come for a little more "can do" spirit. By the way, this is a reversal for me since I too have been very vocal in sounding the alarm and up until recently I was very disappointed in the lack of attention the issue got from many leaders. I don't think the awareness campaign needs any more help and from now on I think those of us who know how to fix this stuff should be fixing things and showing others how to fix things - we should not be beating the dead horse of how serious the problem is. The mainstream media has picked up that campaign and they can't be beat at reporting bad news. I hope my fellow computer geeks can spread some good news so that the people don't panic. We need a little alarm to keep our feet to the fire but we don't need panic. My humble opinion is that we will have problems, businesses will fail but mankind will survive and prosper in the new millenneum.
-- Don Harlow (email@example.com), June 24, 1998.
-- Rocky Knolls (firstname.lastname@example.org), June 24, 1998.
Gee, Don, that's great. Now, please tell us about all the embedded systems that you made compliant,too. While you're at it, please tell us why Senators Bennett and Dodd are as *stupid* as others whom you've raked over the coals with your name calling. And why everyone else who doesn't agree with you should also be termed *stupid.*
You don't advance your point very well by resorting to name calling.
-- Rocky Knolls (email@example.com), June 24, 1998.
Rocky, Sorry I got your bowels in an uproar. I apologize if contradicting someone's premises is a personal offense to that individual. I still think the oversimplification of these issues is no longer necessary and anyone involved in developing software should be able to appreciate the points I so poorly made. You mentioned the honorable politicians who have taken up this Y2k campaign and I am not impuning there efforts. It is precisely because we now have that level of attention that I think we should be a little more careful about exagerating the problem. It just isn't necessary Rocky. Ahh, the dreaded embedded systems problem. Actually, I have been involved in that. We have not been able to locate a single problem in our production facilities. We did have an Air conditioning system that might have been a problem because it once had a computerized contol module. However, we replaced that control module(for non-y2k reasons) with hard-wired dials - by the way if that fixes an embedded system problem on an air conditioner for $xx can we now assume that all the embedded systems in air conditioners can be fixed for the same cost? More on embedded systems - like you Rocky, I have been evangelical in telling the Y2k story to friends and relatives. One of my successes is a friend who is president of construction company. The company has a great deal of heavy equipment and this company president tells me that he now has written assurance from all of his equipment manufacturers that his equipment is compliant. I don't tell you that to imply there is no embedded system problem but to illustrate that people in responsible positions are taking action and many of the people taking action are pleasantly surprised to find that there are few problems in their area. Embedded systems are a serious problem but there are serious people working on it and there efforts are belittled by some of the soothsayers that I may have inadvertently offended. Name calling? If I am name calling then what are you people doing when you imply that most of my peers will sit around on their hands and watch their life's work meltdown and society with it. Will there be problems? Yes. Are all of the people who are working on this problem as incompetent as some of you seem to believe? No. Now let's get back to storing dried food and bullets in the bomb shelters we built in the 60's (oops 1960's.)
-- Don Harlow (firstname.lastname@example.org), June 25, 1998.
Thanks, Don. I'm with you. I think we can do better. Let's not give up the fight yet.
-- Sam Loy (email@example.com), June 25, 1998.
I certainly agree that many of the Y2k problems will be fixed. I am sure we will fix those in my own relatively small company. But I think that some respondents are not adequately considering the interconnectedness of the whole system and society. I think it is highly likely that the FAA will not have air traffic control working for at least several months of 2000. I think a few of GM's suppliers will not be able to deliver parts to the GM factories (we are seeing right now what happens when one small part of that system goes out on strike). I think some government payments will get disrupted. I think a lot of businesses in other countries will be seriously disrupted. Each of these, and all the unforeseeable additional glitches, will be drags on the economy, with ripples out to other companies, causing at least a significant increase in unemployment, and losses in the stock market, and possibly great economic dislocation.
Let's face it, the situation is unique, and no one knows what will actually happen. So let's try to share experiences, insights and opinions without implying that anyone is stupid to have drawn different conclusions than you have. I am sure that in some organizations the correction of the problems has been relatively easy (it is in mine). But we had all the source code, and relatively few systems, and we don't do electronic transactions with any company other than our payroll service provider. I avoid assuming that everyone else has such an easy time of it.
I consider that some of the people who are somewhat alarmist, such as Yourdon, Yardeni and de Jager, have much greater personal information on the extent of the problem than I do, in that they are in direct contact with dozens or hundreds of companies and organizations, and that they are privy to problems about which they cannot speak directly under confidentiality terms of their consulting contracts. And the messages from these guys have been getting more alarmist, not less alarmist, as time goes by and they have more contacts and input from inside sources.
I e-mailed Ed Yourdon directly some months ago making the same argument of a previous writer on this question, that there shouldn't be scope creep in Y2K remediation. His reply to me was as follows:
>>>I agree with just about everything you've said, except for your assumption that the "statement of work" for Y2K projects will be stable. At first glance, that would seem to be so; but there are at least two reasons why it's more complex than it would seem.
First, the number of systems that have to be remediated has been badly underestimated in most cases. The Federal Government has doubled its estimate of the overall cost of Y2K within the past year, which obviously implies that they under-scoped and under-estimated the size of the problem at the beginning. Even more amazing is California: their initial budget estimate, state-wide, for Y2K was $50; by Feb 1998 it was up to $840 million. The reason for this, of course, is that the initial estimates usually include only the mainframe applications. Then someone says, "But what about the PC's?" And someone else says, "But what about the embedded systems?" We're gradually learning that the embedded systems cost nearly 4 times as much to fix as the other stuff.
The other reason for the lack of stability is the gradual awareness of the Y2K risks associated with the "supply chain" of customers, suppliers, partners, etc. I can budget, estimate, plan, and manage the Y2K activities associated with my old legacy COBOL programs; but if my hardware vendor, OS vendor, and database vendors are behind schedule on their Y2K efforts, or have lots of bugs in their Y2K-compliant upgrades, that throws a monkey wrench into my project. If the lights go out, the phones are dead, the Internet crashes, the banks are closed, the airports and schools and hospitals close down, chances are that it will be a little tough to keep things going normally within my own system. <<
-- Dan Hunt (firstname.lastname@example.org), June 25, 1998.
My statement stands. YOU are the one who barged in using abusive language and calling anyone who disagrees with you 'stupid.' That....not the remainder of your arrogance......is what raised hackles.
I repeat. Your language and namecalling serves no purpose.
By the way, since you seem to be upset about lack of recognition, I believe that most of us recognize that a lot of people are working on the problem. I simply believe that --- in spite of islands of success --- I've seen nothing to convince me that a general solution is in sight. You may have reinforced that conviction by protesting too much.
-- Rocky Knolls (email@example.com), June 25, 1998.
I don't really care if some two bit construction company is compliant...or even General Motors. I am more concerned about the power companies. They are the key to our survival!
-- Annie (firstname.lastname@example.org), June 25, 1998.
Again I apologize for my lousy choice of words. Actually I am a big fan of all those (even the gloomiest) who have had the talent and made the time to get the message out. I was responding to the original question of whether Y2k can be solved. I thought it might be worthwhile to point out the falacy in some parts of the statement of the problem. Perhaps I am wrong, perhaps the world is coming to an end and perhaps all of us who are trying to prevent that should just give up. Rocky, I appreciate you taking me to the woodshed over my lack of diplomacy, this old programmer has never been blessed with an abundance diplomatic skills. Having said that I still hope that some people will consider that we may have already won the awareness battle and that fighting yesterday's battle could be counterproductive in the face of mounting public concern. Consider, if you will, how long it has taken for the awareness campaign to gain solid footing. Now all the standard negative arguments are well established and they are all valid but they are also mainstream and the momentum is in the direction of beating the drum louder and louder to the same tune. Is it possible that we could oversell the negative aspects to the detriment of all efforts at solution? Have you read any articles in the mainstream press that are not a repeat of the standard stuff that everyone on this website can repeat chapter and verse? Is there something inherently wrong with reporting successes to balance some of the bad news? Have you read the current ComputerWorld (June 22)? Get the hard copy, look at the faces of those Y2k managers do they look scared to you? Union Pacific railroad has been working on solution and they are guardedly optimistic. Nabisco is well on the way to solving their problems. I believe there is a good chance that Ritz cracker can be delivered by rail in the Year 2000 and even Y2k will taste better when it sits on a Ritz! I will remain optimistic if for no other reason than to avoid losing my motivation to continue working on this problem. Finally as to a guaruntee that everything in general will be fixed. Everything won't be fixed but nothing will be fixed "in general." Nothing is ever fixed in general, things are fixed specifically and this work is being decomposed into discrete, manageable parts that can be worked on specifically by people in each area who know more about their specific areas of expertise than any of us who work in different areas. My thanks to those other folks who have expressed some guarded optimism themselves.
-- Don Harlow (email@example.com), June 26, 1998.