Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.
Original article here.
October 7, 1999
In an age of New Math, where feeling good about the answer is more important than getting the right answer, it's no surprise that almost nobody understands the statistical math behind the possibility of Y2K disruptions.
Beyond the dumbed-down-math agendas in public re-education camps... er, we mean public schools, there's also another reason this is overlooked: it's counterintuitive. Few people, when given the numbers, come anywhere close to the right answer.
Test yourself! If Company A has 200 mission-critical systems, and each mission-critical system has a 3% chance of failing on January 1, what is the overall chance of the entire company failing on January 1?
Stop reading and actually ponder this one for a minute. What's your guess on the answer?
Few people arrive at the correct answer: a 99.8% chance of total failure. That means choice "D" -- 100% -- is the correct (rounded) answer.
IT'S NOT OBVIOUS, BUT IT'S TRUE
Even though these answers are not obvious, they are undeniably real. Unfortunately, pointing out the answers (and the math) to most people does little good: they don't accept it because they don't understand it.
For a full description and a handy chart that shows you how this math works, click here for the "Chance of Failure" chart. We built it just for you. You can use the chart to look up the chance of failure in any multi-entity system.
Unfortunately, nobody in Washington will use this mathematical truth to describe Y2K risks: it sounds too pessimistic. But if the President's Year 2000 Council were actually working to assess the current Y2K situation, they would only need to ask two questions of each agency:
Question #1: How many mission-critical systems do you have?
Question #2: What is the chance of each system failing on January 1?
Then they only need to check our chart.
Agencies that respond by saying they have one hundred mission-critical systems and each system has only a one percent chance of failure would seemingly be reporting good news, right? But in fact, this is terrible news because it indicates a 63.4% chance of total failure.
This is precisely how so-called "good news" isn't really that good once you apply logic to the data. For a government agency to be admitting it has a 63.4% chance of outright failure is hardly comforting.
BUT IT'S ACTUALLY MUCH WORSE THAN THAT
Many agencies have far more than a hundred mission-critical systems. Government-wide, the number is reported at approximately 6,000. (It used to be 9,000+ systems before 3,000 systems were taken off the list to boost the appearance of "progress.") What are the odds of all 6,000 mission-critical systems working on January 1? Virtually nill.
But the big numbers start to come into play when considering third-party vendors. How many vendors does Ford rely on to build cars? Estimates have put this as high as 20,000. Obviously, a miracle would have to occur for Ford to not experience supply-line failures. The question, for Ford, is one of making sure the really important vendors will work -- and having contingency plans for the expected failures of the other vendors. For the legal record, we're mentioning Ford only as an example here; we don't know the actual Y2K compliance status of Ford's vendors.
CALCULATING THE NEED FOR PREPAREDNESS
These same mathematics are useful for demonstrating the need to prepare with extra food, water, cash, medicine, heating supplies, and so on. Here's why:
If you knew a nuclear missile was going to strike the United States, but you didn't know where, would you be wise to prepare for it? Would the rest of the country be wise to prepare for it? Or should everybody do nothing because nobody knows where it will hit?
Now consider water. There are over one thousand water treatment facilities in the United States. If each one has a 1% chance of failing, what is the overall chance that at least one facility will fail? The answer, of course, is 100%. (Technically, 99.9957%.) That means somebody will be without water in this country. But none of us knows which area will be hit. We also don't know the duration of the failure in question.
Given this, is it wise to prepare? Remember, there is a near-certain chance that some municipality in the country will be without water for some period of time. But who will it be? And what will be the duration of the failure?
See, there are plenty of knowns and unkowns with Y2K, but not in the way you've been told by others. The unknown is who will be hit. The known factor is that somebody will be hit. The rest is all debatable: should an entire country stockpile extra water to prepare for a near-certain water failure in one city?
To me, this is the best reason not to believe that Y2K is going to be just a "bump in the road."
Very well put... Thank you... EVERYONE NEEDS TO SEE THIS...
I work for a major utility that supplies power and water for about a million people. We have moved all our clocks forward on all of our facilities and everything looks good although we are cutting it a little close on a certain specific control capability (but even that appears to be on track). I do not think that everything is OK everywhere, but I am tired of speculation and non-specifics and foolish conclusions regarding Y2K like this article. I would like to see specific tests and failures that certainly the programming community must be familiar with as these programs are investigated. The biggest threat to the US I see is sabatage and possible foreign supply disruptions and perhaps an economic judgement for a nation that has rejected the commandments of God.
The only thing wrong with this analysis (and the math is right) is that it implies an instantaneous failure, and an instantaneous end of the company, or whatever. That's not true, of course, because many companies have been shut down for a day or even much longer and still survived.
To be accurate, each of these failures has to be permanent, unfixable, and with no adequate workaround within the time frame necessary to avoid a business collapse. Y2K may not work that way.
What it is deadly accurate on, however, is how prevalent Y2K problems will be at some point next year. Virtually everyone will have them, and it's going to be interesting, if somewhat scary, to see if we're nimble enough to deal with them. Three-day snowstorm, my butt.
"The only thing wrong with this analysis (and the math is right) is that it implies an instantaneous failure, and an instantaneous end of the company, or whatever. That's not true, of course, because many companies have been shut down for a day or even much longer and still survived. To be accurate, each of these failures has to be permanent, unfixable, and with no adequate workaround within the time frame necessary to avoid a business collapse. Y2K may not work that way. What it is deadly accurate on, however, is how prevalent Y2K problems will be at some point next year. Virtually everyone will have them, and it's going to be interesting, if somewhat scary, to see if we're nimble enough to deal with them. Three-day snowstorm, my butt."
But it's the instantaneous threat that is the real threat. A sreies of breakdowns can be dealt with like the fiscal year rollovers in 1999. If there is no instantaneous threat, fix-on failure won't be Armaggedon. The problem with the lights and/or water going out at the stroke of midnight is that civil unrest will occur immediately, the people instantly and correctly discerning that they have been betrayed. That is the problem. If everyone stays calm, the programmers, et al will be able to rig systems to work, and therefore, the problem will be solved.
I work for a major utility that supplies power and water for about a million people. We have moved all our clocks forward on all of our facilities and everything looks good although we are cutting it a little close on a certain specific control capability (but even that appears to be on track).
Do you know if other areas have been checked, I mean besides where you can physically change the date. For example, date sensitive chips, PLC controllers, etc.
I do not think that everything is OK everywhere, but I am tired of speculation and non-specifics and foolish conclusions regarding Y2K like this article.
My intent is to get people to start at least thinking about the possibility that everything isn't going to be okey-dokey...and then to prepare. In other words, to maybe save a life or two.
I would like to see specific tests and failures that certainly the programming community must be familiar with as these programs are investigated.
So would I. That's one of the problems: Nobody is really forthcoming with the tests they've done. Businesses are naturally hesitant to share the results of their surveys and/or testing, partly because of adverse results, partly because of fear of competitors knowing their systems, and partly because they don't want to spook their customer.
The biggest threat to the US I see is sabatage and possible foreign supply disruptions and perhaps an economic judgement for a nation that has rejected the commandments of God.
Then what's the difference if one prepares for Y2K or if one prepares for one of these events?
" The only thing wrong with this analysis (and the math is right) is that it implies an instantaneous failure, and an instantaneous end of the company, or whatever. That's not true, of course, because many companies have been shut down for a day or even much longer and still survived. To be accurate, each of these failures has to be permanent, unfixable, and with no adequate workaround within the time frame necessary to avoid a business collapse. Y2K may not work that way. What it is deadly accurate on, however, is how prevalent Y2K problems will be at some point next year. Virtually everyone will have them, and it's going to be interesting, if somewhat scary, to see if we're nimble enough to deal with them. Three-day snowstorm, my butt."
The problem is that the utilities might just go away at the stroke of midnight. See the earlier thread today about delivering pizza. You gonna expect a programmer or a Telephone/Electric Co. worker to go out and replace chips while most people are dying of thirst, hungry and cold? Are people going to riot? I'm really not happy about this. This is the Cindarella syndrome. She must leave the ball before the stroke of midnight, or she loses all.
My point is that if nothng much happens at 12:01, Dec 31, 1999, then nothing much more of consequence will. Don't drag your pride into the coming year.
The fundamental problem of the Y2K computer defect is that failures will happen all at once, not that they will simply happen.
Do you know if other areas have been checked, I mean besides where you can physically change the date. For example, date sensitive chips, PLC controllers, etc.
Your questions implies that there are systems that are aware of the current date, but have no way to set the current date. This is impossible. Any system that is susceptable to Y2K problems must have some way of setting the date. Even if they normally read the current date from the GPS system (or some other automated system), it is always possible to "fake" the source of the date information for the purposes of testing.
Keep well.
John
More FUD.
If the analysis in this article was valid, then every time a mission-critical system failed, for any reason, the company would fail. This is (obviously) nonsense.
I was involved in fixing a "dying" credit card processing system for a large hospitality-industry company (the most "mission-critical" system they had, especially since amost 100% of the revenues came from credit card transactions). For several months, we struggled with almost constant problems. It meant lots of long days on the part of the technical people involved, lots of screaming from management and a small number of very annoyed customers. But the vast majority of customers had no idea that there were any problems at all.
For any of these doom&gloom predictions to come true, the people responsible for the systems would have to do nothing to fix and/or work around the problems, and that just isn't going to happen.
Keep well.
John
The problem is that the utilities might just go away at the stroke of midnight.
It's my understanding that at least some of the power companies are going to "stagger" the clocks on their embedded systems, so that (for example) each generator will have a different notion of the current time. That way, even if there is a problem that causes the generators to shut down, it will only affect one generator at a time (which happens routinely and doesn't result in any power failures) instead of shutting down all the generators at the same time (which would result in at least brief power disruptions).
Note that they don't expect this to be necessary. But this is a simple, foolproof and effective way of dealing with any problems they might have missed during testing.
Keep well.
John
"I work for a major utility that supplies power and water for about a million people. We have moved all our clocks forward on all of our facilities and everything looks good although we are cutting it a little close on a certain specific control capability (but even that appears to be on track)."
Your employer isn't a manufacturer dependent on large numbers of components from large numbers of third-party suppliers to continue daily operations.
Your number of suppliers for critical daily-operation items, ranging from chlorination chemicals to coal to generate electricity, are relatively few in number - and U.S.-based. In no way is a power utility or a water company comparable to an auto company or a computer company that depends daily on getting in today's shipment of speedometers or hard drives - often made in foreign countries with notorious Y2K problems in Asia.
We were warned about this "supply chain risk" at a lawyer's conference in Winston-Salem, North Carolina in July - and told that it's unavoidable and unpredictable.
Let's take payroll as an example of a mission-critical system. It may effectively break on January 1, but if it doesn't have to process until the 10th, it's not an instantaneous failure. The company doesn't go out of business January 1. If the company can make incorrect payments, but close enough to keep the wheels on, they have bought some more time on that system.
But some failures will be instantaneous, many perhaps in embedded systems that were ignored or wrongly assumed to be compliant for some reason. If the big mainframes in KC that switch the nation's railroads malfunctions, that will be pretty close to rollover. Same with Air Traffic Control. And power, of course. Any of these can quickly prevent a company from effectively fixing its own failures, and can even prevent a fully compliant company from functioning normally.
That's the real danger of Y2K; that a failure in one industry cascades into the other industries that depend on it for vital business functions. If Y2K avoids widespread instant failures, we will work our way out of it, with only a recession. If we can't avoid the instant failures, it spirals downward quickly and nobody can predict where the bottom is.
Your questions implies that there are systems that are aware of the current date, but have no way to set the current date. This is impossible. Any system that is susceptable to Y2K problems must have some way of setting the date. Even if they normally read the current date from the GPS system (or some other automated system), it is always possible to "fake" the source of the date information for the purposes of testing
You don't have to believe me, but if you would like, you can read all about Y2K and the embedded chip risk at The Institution of Electrical Engineers Y2K risk website. Especially interesting is the process cycles treatise.
Ya know, it's a relatively easy thing to do research on the embedded chip problem, especially if you have internet access, which you plainly do.
If the analysis in this article was valid, then every time a mission-critical system failed, for any reason, the company would fail. This is (obviously) nonsense.
Ummm...if someone designates something as "mission critical" then that implies that if that component doesn't work, then it's critical (life threatening) to the business. Now you would be correct if they designated something as mission critical when it really wasn't.
I was involved in fixing a "dying" credit card processing system for a large hospitality-industry company (the most "mission-critical" system they had, especially since amost 100% of the revenues came from credit card transactions). For several months, we struggled with almost constant problems. It meant lots of long days on the part of the technical people involved, lots of screaming from management and a small number of very annoyed customers. But the vast majority of customers had no idea that there were any problems at all.
Well Johnny B, in your wonderful example above, it doesn't sound like you had a total failure...more like a lot of problems. Obviously the system didn't "fail".
Ya know, it's a relatively easy thing to do research on the embedded chip problem, especially if you have internet access, which you plainly do.
I've developed embedded systems.
Most embedded systems don't know or care what the date is. Most of the rest will work correctly (embedded systems rarely store dates in a format that is susceptable to Y2K problems). Of those that do have Y2K bugs, most are not critical. Of the remainder, most will have been caught and corrected (or at least worked around) in Y2K testing.
Of course, some will fail. But then, some of them have failed every day since embedded systems were first created.
Keep well.
John
I've developed embedded systems. Most embedded systems don't know or care what the date is. Most of the rest will work correctly (embedded systems rarely store dates in a format that is susceptable to Y2K problems). Of those that do have Y2K bugs, most are not critical. Of the remainder, most will have been caught and corrected (or at least worked around) in Y2K testing. Of course, some will fail. But then, some of them have failed every day since embedded systems were first created.
I don't know how to take you. Where do you get your information? Did you actually read the information on the IEEE website? Do you know what the IEEE is? This organization estimates that between 6% and 7% of all embedded chips will fail ALL AT ONCE on the date rollover. They ARE NOT Y2K alarmists. They are engineers and scientists who develop STANDARDS for the electronic and computer industries.
Like I said, you don't have to believe anything I say, but when a fireman tells you your house is going to burn down, I would believe him.
I don't know how to take you. Where do you get your information? Did you actually read the information on the IEEE website? Do you know what the IEEE is? This organization estimates that between 6% and 7% of all embedded chips will fail ALL AT ONCE on the date rollover.
Please post the URL for this information if you have it. But, since we're throwing around "expert's" opinions, here's one (which is referenced on the IEEE web page) that you might find interesting:
Update on Embedded Systems Failures--My notes on embedded systems are at http://www.Y2K-Status.Org/Embedded/ now and the older page you mention is still available. The average situation seems to be that about 3-5% of subsystems with embedded chips will have Y2K non-compliance issues - that is, they will fail specific Y2K tests. However, apparently, only 3-5% of these, in turn, will cause the systems they are in to fail. If such systems are left to fail, whether that system failure causes a serious business or functional failure will depend on the amount of preparation and contingency planning. Patrick Bossert has been measuring many systems and is finding 5% and 3%. Senator Bennett reported recently 2% and 2%, which he attributes to Intel. GartnerGroup reports 1 of 100,000 of free-standing chips will fail. The difference between a free-standing chip and a subsystem is the number of chips, their interactions, and the likelihood that a systemic Y2K problem is not found. Few expect serious problems with missed embedded systems, but they will spike around 2000.00 and are still poorly documented. I am constantly adding new special studies, and recommend http://www.Y2K-Status.Org/KeyReports.htm to those concerned about the global aspects of the problem.Regards,
Richard Collins, Director
http://www.TheInternetFoundation.Org
Year 2000 Project - http://www.Y2K-Status.Org FCC Releases Y2K Assessment of
According to this "expert", the actual failure rate for embedded systems is less than one quarter of one percent (5% x 5%), and it is not reasonable to assume that all of the failures will result in serious business or functional failures.
Since one can find "experts" who support the entire range of opinions (from "absolutely nothing will happen" to "civilization is going to collapse"), I tend to disregard any such statements. I'm basing my opinions on my personal, extensive experience, and the experiences of everyone I know in the industry. I don't know anyone who is actively involved in software development who is seriously worried about Y2K.
FYI, 18 months ago, I would have qualified as a "doomer". I was seriously considering finding a place to hide for a couple of months on either side of Jan 1. However, since then I've been involved in several major Y2K remediation projects, and I've been comparing notes with all of the other programmers I know. None of us have found anything to be worried about. The problems were smaller than we expected, and in general, wouldn't have caused major problems (the kind that would put the company out of business) even if they had not been fixed before they occured.
The "doomers" have been consistently wrong about the effects of Y2K bugs for every deadline we've reached so far. Nothing significant happened on 1/1/99, 4/9/99, 9/9/99, or the first of any month that began a fiscal year for some big business or government agency. I have no doubt that they'll continue to be wrong throughout 2000.
Keep well.
John
So, what's your point in all this? What have you proved, and to whom? Why do the kooks and loons matter to you? Why are you so passionate in your efforts to prove everybody wrong? Why do you spend so much time hanging out on threads where you disagree with everybody?
Well Johnny B, in your wonderful example above, it doesn't sound like you had a total failure...more like a lot of problems. Obviously the system didn't "fail".
Obviously, the system did fail, by any reasonable definition of the word.
Do you really think that Y2K will cause computers to just lock up solid??? The vast majority of Y2K bugs are exactly the same nature as the example I described above: a running system that's producing garbage. And, while that's a huge pain to those of us who have to keep such systems working, we have a lot of experience doing just that (keeping broken systems running).
You're so predictable you're hilarious!
So, what's your point in all this?
I know better than to think I'll "convert" any of the doomers. But I'm hoping that I may calm the unreasonable fears of the majority of the readers who don't take part in these discussions, and don't know enough about the problem to have formed their own reasoned opinion.
I seriously think that the effects of Y2K panic will outweigh any real Y2K problems. And if enough people start panicing (closing out their accounts and the like) Y2K may become a self-fufilling prophesy, even if their isn't a single technical glitch.
That's my point.
Johnny B. saves the world. Your mama and daddy will be proud. Not to mention the kids on your street. You'll be quite the hero.
Johnny B. saves the world. Your mama and daddy will be proud. Not to mention the kids on your street. You'll be quite the hero.
Since you're now resorting to ad hominem attacks, I guess that means you can't refute the comments I've made.
Have a nice day.
I am.
Hi John,
Have you read this? International Monitoring Senate Testimony of 10/13/1999
Just more doomer lies?
So, if Y2K is bad, we will send everyone over to your house!
Oh, whats that? You don't have any food? Guess you will be livin it up at the local FEMA shelter with 1500 of your closest friend! What fun.
According to this "expert", the actual failure rate for embedded systems is less than one quarter of one percent (5% x 5%), and it is not reasonable to assume that all of the failures will result in serious business or functional failures.
You know, this is really my whole point in all of this. You cite an "expert" that believe that there WILL be this many failures. I cite an expert that believes there will be more failures. There's a lot of experts, but not one expert says there will be NO failures. Now given that there will be failures, and given that nobody actually knows how many failures there's going to be, wouldn't the smart money be to prepare for the worst?
Well Johnny B, in your wonderful example above, it doesn't sound like you had a total failure...more like a lot of problems. Obviously the system didn't "fail".
Obviously, the system did fail, by any reasonable definition of the word
If you're going to try to convince me that "fail" means "work with problems" with some type of Clintonian word games, then I ain't falling for it.
When you fail a test, do you pass it with a poor grade?
There's a lot of experts, but not one expert says there will be NO failures.
And neither have I. But there's a huge difference between "no failures" and "catastrophe" (to use "Jethro Tull's" term). What I am anticipating are numerous small failures (of the annoying or amusing kind, depending on whether they happen to you or some other bloke), and some significant problems. I know you don't want to believe this, but it's not at all unusual for computer systems to have significant problems. Those of us in the business have a lot of experience dealing with these issues.
If you're going to try to convince me that "fail" means "work with problems" with some type of Clintonian word games, then I ain't falling for it.
In every Y2K testing standard I've seen (quite a few), a "failure" is considered any erroroneous action or result brought about by incorrect date calculations. If you really think that the term "Y2K failure" means "the computer won't start at all" then you really don't understand the problem at all.
One thing I've noticed (again, based on my first-hand experience) is that businesses aren't making any distinction between catastrophic failures and cosmetic flaws. The standards that I (and everyone else I know of) have been working under for Y2K remediation work was total, absolute perfection. The most trivial cosmetic flaw was enough to prevent "Y2K compliance". This quest for perfection may be noble, but it's hardly necessary to insure that the lights will stay on and the store shelves will stay stocked. "Good enough" is a long way from "perfect".
You keep basing your opinions on those media reports that support that opinion, and I'll keep basing my opinions on personal, first-hand experience.
I know you don't want to believe this, but it's not at all unusual for computer systems to have significant problems. Those of us in the business have a lot of experience dealing with these issues
Maybe you don't know my background...I've posted it enough to feel like I'm being redundant, but here it goes again:
I joined the USAF in 1982. My dad had been into computers since the 1960's. At that time, he was the systems manager at a large paper mill in our area. We were one of the first people in our town to have a TRS-80. He recommended that I join the Air Force as a computer repair tech. I did.
I went through 9 months of tech training in Mississipi. The rest of the time, I was a computer repair tech at Ellsworth Air Force base in South Dakota. I serviced communication computers that were used to relay alert messages between SAC headquarters and the nuclear missile capsules (where the officers hung out and launched the nukes if needed.) My security clearance was higher than my IQ.
I got out in 1986. For 5 years I worked in data communications for the banking industry, mainly in setting up and repairing ATM's. For fun, I learned how to program in Pascal (I already knew BASIC) on the new IBM PC's that were becoming all the rage of business.
Over the years, I held various positions...network administration, help desk management, IS management. My present title is Director Of Information Systems for two corporations (common owners) in Michigan. Both companies are highly automotive industry related. One makes interior seat parts, the other chemically coats and/or paints automotive parts, expecially bumpers, for major manufacturers.
I've worked on Unix systems, Novell and Windows NT networks. I'm an MCSE. I passed all the required tests within a two month period in early 1998.
I love the field...computers are about my only hobby. I'm very good at what I do.
I don't tell you all this to try and impress you. I'm telling you this so you can understand that I approach this from a much broader perspective then you might imagine.
Sounds similar to what I did in the USAF. I was a 306X1 from '74 to '82
OK, thanks for the resume. :-)
To return the favor, I've been programming for 20 years, mostly C, but also several assembly languages, BASIC, several database systems (Informix, Sybase, Progress). I've designed and developed accounting systems, realtime database systems for the (at the time) largest options trading company in the world, and embedded systems for commercial airliners, among other things. I spent a considerable chunk of this year Y2K testing the reservation system for a major hotel chain (name starts with an "H"; you've heard of them) that I helped develop.
Now that we got that out of the way... Do you (and those coworkers and associates of yours who actually know what they're takling about) really expect things to fall apart? Everyone I know in the field (the people actually doing the testing and fixing, not necessarily the PHBs) expect Y2K to be pretty much a non-event.
One of the things that annoy me about the "doomers" is that they seem to assume that if a computer system fails, everyone will just give up and go home (to sit in the dark, no doubt). I've been in the middle of numerous computer system failures, and that just doesn't happen!
I've never said that "preparing" was a bad thing. In fact, Y2K supplies came in handy here in Chicago a month or two ago when we had a big power failure downtown (no, it wasn't Y2K related). But I really am more worried about people overreacting to Y2K than I am to Y2K itself. I expect absolutely everyone to top off their gas tanks, and stock up on food, starting about 12/30/99. If retailers haven't prepared for this spike in purchases, there'll be spot shortages for a few days (but not long term, because the average consumption won't be changing).
But let's not forget that this is the most thrououghly anticipated, researched, reviewed and prepared for "emergency" in the history of the world.
BTW, a year ago I was a doomer myself. Keep in mind that there have been Y2K failures anticipated throughout this year (1/1/99, 4/9/99, 10/1/99, etc.). The fact that there haven't been any significant problems so far has convinced me that the problem has been overblown and/or largely repaired. The "doomers" just keep moving their predictions into the future, saying, "you just wait for the next drop-dead date! That's going to be the big one!"
My security clearance was higher than my IQ.
Somehow, I'm not surprised.
DouglasKC, in order for an embedded chip to fail, whatever process is using it will have to actually make a procedural call to access the date. The Navy has a system for message processing that is so dumb, it never makes a date call. The hardware and operating systems are theoretically non-Y2K-compliant. However, because there's no date calls, the system runs like a champ in rollover and Battle Group Systems Integration Testing.
Somehow, the idea that a system will make a date call without there being any way to reset the date strikes me as being REAL dumb. Resetting the power will take the chip back to its earliest date. Take my car (please). When the battery is disconnected (or goes dead, as has happened several times), all the chips lose all the data. I figure my car will think it is mid-1983 (when it was built), assuming that it cares what the date is.
Johnny B,
I see some of your experiences are matching mine. Our Y2K projects had to be absolutely perfect or they were deemed non-compliant. Functionally things worked our first time but some silly headers on some reports (absolutely no connection to the processing at all) showed a date of 100/01/01 instead of 00/01/01. Had to fix those trivial things and then back through the testing loop again. Never has any project we worked on been 'watched' as closely as this one and it's getting annoying. Software breaks, or fails, all the time and then we programmers go out and fix it. Y2K is just another bug, only this time the world will be watching.
As far as the math that is involved (at the start of this thread) I believe the author is working on a flawed premise. Within the same company things have a way of being done the same throughout the company no matter what department you are in. So that 1% chance of failure with over 100 systems may truely be a 1% chance of failure since those systems are dependant and may share code (especially something as trivial and global as date/time calls) so that fixing it in one place may actually fix all 100 systems. If the first system in the chain works then there is a much higher chance that each suceeding link in the chain will work also since they are more than likely based on very similar code.
Here is another log to add to the fire. The Social Security Administration claims to be Y2K compliant concerning their major systems. However, there is one fly in the ointment. SS checks are printed by the Treasury Department. According to a source two weeks ago, the printing facility at Treasury is not and will not be ready to print SSI checks. So much for accountability. The finger pointing will be a source of amusement to watch as things unfold.
Besides the condescending tone of the author (What if I disagree with these numbers? We recommend the book, "Everyday Math for Dummies" by Charles Seiter,.....), there are several problems with this article. Namely: the application of the probability equations, the selection of the chance of failure percentage, and the assumptions about the designation of "Mission Critical Systems".
The math behind it is correct, but it is not a correct representation of the true system. The expression "100(1-(1-.03)200 )= 99.7739" correctly represents a series system of 200 components, where a critical failure of one component results in a critical failure of the system. However, the probability of failure of a system depends upon several factors: the configuration, (series, parallel, or combination), effect of the failure on the system, and redundancies, just to name a few. And, contrary to the author's opinion, accurately applying a probability of failure and determining the effect of the failure is not a simple task.
One way to determine the probability-effect of failure is to complete a Failure Mode and Effects Analysis (FMEA). FMEA is a procedure by which potential failure modes in a technical system are analyzed. An FMEA can be extended to perform what is called Failure Modes, Effects and Criticality Analysis (FMECA). In a FMECA, each failure mode identified is ranked according to the combined influence of its likelihood of occurrence and the severity of its consequences.
An accurate FMEC analyzes the occurrence and probability of each failure mode and the criticality level of the failure effects are analyzed. Failure criticality can fall into one of four categories (from: An Introduction to Reliability and Maintainability Engineering, Charles E. Ebeling):
So what, specifically, is wrong with the author's thesis? For starters, the correct answer to the initial question is not "D". The correct answer is, "Not enough information"!
In order to accurately answer the question, one would need to know more about the systems. "System" implies an interconnection or interaction. But, in reality the author does not know how the systems are interrelated. Since we have no information, I will hazard a guess: The systems have little or no interconnection. For example, within an agency or business there may be two critical systems. One is for purchasing and the other runs the production function. But, there is no interrelation between the two. Therefore, using a probability equation that assumes a serial relationship (as the author does) is not valid.
The next question is, Just what is a "Mission Critical System."? In the example from the article we are told that the company in question has 200 Mission Critical Systems. Just how was that determination made? Was it made in accordance with some standard, such as IEEE or MIL-HDBK-217 or Bellcore? Probably not. In fact, they were probably made using the same method that our friendly House of Representative did.
Just look at the House web site, and you will see some of the systems that they have designated as "Mission Critical". Here are a few examples:
Photography (FOTO) - The FOTO application tracks the use of photographic services by Members, Committees, and other House offices.
Recording Studio (Studio) - The Recording Studio application tracks the use of the House radio and video facilities by Member, Committee, and other House offices.
Parking (POPS) - The POPS application tracks parking permits issued by the House Garage and Parking Security Office (Parking Office)
Hmmmmm.....the end of civilization as we know it? Because our Representatives can't track their parking passes? I'm not buying it!
But, while were are on the subject of Buying, lets go back to the authors " Chance of Failure Chart". It is sponsored by Arial Marketing, Inc....what do you supposed they are marketing?
PS: Do you supposed that the "KC" in our names has something to do with the fact that our resumes are so similar? ;)
SAC will be back!
"As far as the math that is involved (at the start of this thread) I believe the author is working on a flawed premise.
I agree...see my post above.
Now that we got that out of the way... Do you (and those coworkers and associates of yours who actually know what they're takling about) really expect things to fall apart? Everyone I know in the field (the people actually doing the testing and fixing, not necessarily the PHBs) expect Y2K to be pretty much a non-event.
Here's what I expect: A 6 to 12 month of real economic hardship, precipitated by Y2K failures.
Why? I currently work in the automotive industry. I know that just one supplier plant going down (not able to ship product) because of a strike can bring down a GM or a Ford or a Daimler Chrysler and their thousands of suppliers. Our plants were idled for two months in the summer of 1998 because the workers at one plant that supplied parts for GM decided to go on strike. Another way to look at it: It's takes thousands of independent companies to make 1 automobile.
I think the odds are overwhelming that not all the companies actually checked their Y2K readiness, especially at the smaller companies. I think it's reasonable to assume that some company presidents think of Y2K as just a lot of hype, or don't even think about it at all. I think that business is made up of about the percentage of people that post here: Many just don't think of it as a problem at all. I think that it's almost a given that multiple plants are going to fail (NOT BE ABLE TO SHIP PRODUCT) in just the automotive industry alone.
Now keep this in mind: The automotive industry isn't that unique. All industries have tight supply chains. Nobody wants to keep excess inventory sitting around. In fact, thats one of the holy grails in manufacturing: Excess inventory = dollars wasted.
This isn't even taking into account the non-awareness of other countries, on which many many US companies recieve goods and services.
Yep, I think the odds are pretty darn good that multiple industries are going to experience multiple failures that are going to manifest as economic hardship.
Our plants were idled for two months in the summer of 1998 because the workers at one plant that supplied parts for GM decided to go on strike.
One big difference is that in a strike, people are purposely trying to disrupt production. In any Y2K problem, they will be actively working to minimize any disruptions. Also, while that may have been painful for the people involved, the vast majority of people never even noticed that strike.
Recently, we've had a major hurricane on the east coast, and a major earthquake in Taiwan. Both were genuine disasters. Both resulted in significant loss of productivity. But neither one has had a significant effect on the world economy, and neither had much (if any) effect on the global infrastructure.
Businesses shut their doors every day, and I'm sure that for some of them, Y2K will be the straw that breaks the camel's back. If the automobile industry (to use your example) is so fragile that they won't be able to work around whatever bugs they've missed, even when everyone involved is doing their best to fix the problems, then I guess we will have problem. Time will tell.
What kind of work do you do for the auto industry? If you're a programmer with first-hand knowledge of the Y2K remediation effort, perhaps you can fill us in on the problems that have already been discovered and fixed, and the number of problems that have already bitten your company. I'm sure your systems are already routinely dealing with post-Y2K data. If your systems (or any of the systems that interface with yours) had serious Y2K problems, you should already be having problems. Again, I'm genuinely interested in hearing first-hand experiences with Y2K issues, good or bad.
Keep well.
John
But, while were are on the subject of Buying, lets go back to the authors " Chance of Failure Chart". It is sponsored by Arial Marketing, Inc....what do you supposed they are marketing?
I’ll answer my own question.
From the Y2K Newswire site: One Year Basic Food Unit + Y2K Newswire Membership for $569.
What kind of work do you do for the auto industry? If you're a programmer with first-hand knowledge of the Y2K remediation effort, perhaps you can fill us in on the problems that have already been discovered and fixed, and the number of problems that have already bitten your company. I'm sure your systems are already routinely dealing with post-Y2K data. If your systems (or any of the systems that interface with yours) had serious Y2K problems, you should already be having problems. Again, I'm genuinely interested in hearing first-hand experiences with Y2K issues, good or bad.
I'm the Director of Is. I was fairly lucky here as far as Y2K goes. I replaced older PC's that had BIOS problems. Not all companies will do that. We're running NT 4.0 SP4 on servers I installed last Feb, so not a whole lot of concern data wise. Other companies won't replace anything. PWhat kind of work do you do for the auto industry? If you're a programmer with first-hand knowledge of the Y2K remediation effort, perhaps you can fill us in on the problems that have already been discovered and fixed, and the number of problems that have already bitten your company. I'm sure your systems are already routinely dealing with post-Y2K data. If your systems (or any of the systems that interface with yours) had serious Y2K problems, you should already be having problems. Again, I'm genuinely interested in hearing first-hand experiences with Y2K issues, good or bad. Our busines software will be okay, but other companies won't bother to check. The phone system was not compliant...but that had been taken care of before I got here (about a year now). Pretty simple operation, basically mechanical build at the one location. The most worrisome thing is the suppliers. We have compliance letters from all of them, stating that they don't expect problems, but we have between 75 and 100 suppliers at any one time for our products, which is not a whole heck of a lot. There's no way to verify that they're really okay...and basically that's all anybody is doing, going on promises. Our fix is to stock up on enough inventory to last us three weeks, which is a huge amount.
The last company I worked for, I was the IS Manager/DB programmer. Part of my job was to maintain and modify a custom written Foxpro app that ran the business. I know that it had a billion places where the date would cause problems if it rolled over...because I worked with the source code every day. I managed to get that replaced just before I left there. The parent company of the company I worked at was quite a bit larger ($250 million company). They had an older mainframe system that ran BMACS, a Unisys product. It was not compliant and would have required a minimum of $500,000 to "fix". The place was huge and had hundreds of production machines,controllers, multiple buildings, alarm systems, etc. A Y2K nightmare. I had met their IS director quite a few times. Y2K remediation literally drove him nuts. He had a nervous breakdown because the board and the President of the company refused to allocate significant resources to fix the problems. He checked into a hospital for a month and never returned to work. That's when I looked for another job. That company probably isn't going to make it. They're going to be lucky if their doors unlock Jan. 1.
Another interesting aside. We we're looking at a software demo yesterday for some software we're considering purchasing. One of my questions had to do with Y2K compliance. They said everything was okay. During the demo though, I noticed that all the dates they were using as examples were two digit year fields and he never had entered anything past 1999. For example, he would do a sales quote and never put the date expected past "99". I insisted that he put a date in that ended with the year "00". He did so reluctantly. When he did, the field interpeted the "00" as the end part of a time string, it dutifully finished it as "01:02:00", in time format, instead of "01/02/00, in date format". He said, "Hmmm...I've never seen that before, and changed it to 99 and plowed on.
What's amazing about this story is the fact that so late in the game this company would have new software with glitches in it and actually try to sell software that had "glitches" in it.
There are countless other examples of "glitches" that have shown up in government systems, other business systems, etc.., all BEFORE the date change. These have been well documented in the media and in this forum by us "doomers"
Johnny, believe it or not, I used to be a "polly" like you. I thought that the whole problem was overblown because I was like you. I did a good job, I made sure my shit always worked and I assumed that everyone else would too. But what really turned me to the dark side was this: As part of my Y2K prep work, I had to get a letter of compliance from our local power utility. This was March of this year. It was a nice little pamplet, I'm looking at it right now. They did not start doing anything at all until September of 1998. They had designated 24 items as "critical" to the successful delivery of power to their customers. As of March 1999, they had "readied" (whatever that means) 75% of these critical items. One of these items was the replacement of the main control panel at the main generating station. They were expected to be down (buying power from other utilities) for two weeks. They were down for two months. As of July, they announced that they're now 100% Y2K ready. HOWEVER, and I quote:
"Although the Board of Light & Power's Y2K preparations are completed, it is not possible for us to know with absolute certainty that all Y2K issues will have been identified and addressed prior to January 1, 2000. This is due, in large part, to the fact that the Board of Light & Power relies upon many other business entities and systems whose Y2K readiness may affect our day-to-day operations. These include, for example, other electric utilities, electric transmission systems, and communications providers."
Now if this doesn't make you prepare for the worst, you're crazy in my opinion. Even if you truly believe that they could have finished all Y2K replacements, testing and remediation in a 9 month period then keep this in mind: a power company, the only one I've ever bothered to check, DOESN'T REALLY KNOW IF THEY'RE GOING TO BE PRODUCING POWER ON JAN.1, 2000.
This isn't meant as a criticism of you, but more like a compliment. I think your problem is that because you've been on the front lines and have been doing a good job getting stuff ready, you are assuming that every company and every industry has bothered to put Y2K on the front lines and treat it as a serious threat. I hope I'm wrong, but I've become convinced that people are people and will take the easy way out. There's a large percentage, worldwide, who will do absolutely nothing until they see what happens in January. Yes, maybe they will all be little glitches that will be quickly solved. I hope so. But the odds are high that there will be many *failures* (inability to ship product) across multiple industries. Consider the effect this would have on the economy.
I'll take the silence on my post regarding the probability and effects of Y2K failures as agreement. Interesting that the claims in the article that started this thread are questionable at best, and yet we continue the discussion
There is no doubt that substantial progress has been made. Nobody disputes that. There is also no doubt that many companies won't beat the deadline, especially since some aren't even trying!
The cost of Y2K litigation in this country has been estimated at one trillion dollars. That sounds like more than glitches to me.
Thanks for your imput - I agree with you that some people here at FR who should know better (programmers, IT staff) are still saying Y2K will be nothing. I'm really suprised at this, because on every other topic, they feel that the government is lying. So, why on earth would they think that the government is telling the truth now?? I don't get it.
Hmmmmm.....the end of civilization as we know it? Because our Representatives can't track their parking passes? I'm not buying it
I think that the number of people who think that Y2K is going to be the end of the world as we know it is relatively small. I don't believe that to be the case.
As far as the original article, I believe the intent was to illustrate exactly how percentages translate to real world information. The author is perfectly correct in what he states as long as long as critical systems actually *are* critical and as long as failure means failure (ability to supply product such as water, electricity, etc.). It's not his problem is someone designates something as critical when it's not.
BTW, what does the KC stand for? It's my last two initials. Where were you stationed?
. That way, even if there is a problem that causes the generators to shut down, it will only affect one generator at a time (which happens routinely and doesn't result in any power failures) instead of shutting down all the generators at the same time (which would result in at least brief power disruptions).
Electricity doesn't work like this. Electricity works on demand. If the electric utilities stagger the electric generation at their disposal and it falls below the power required, then you will have a severe brown out and or failure of the grid. This would also have pretty tough aspects on things like computers that run in a very narrow voltage range, they don't like brown outs. That's why during periods of peak loads your UPS may come on even when the lights remain on. You'er in a brown out or low voltage situation and the UPS kicks in to carry the load.
Or do you mean they will all be running but with different time clocks?
KC, as in KC-135 or KC-10. I was stationed at Chanute, Blytheville and March.
I agree that the author is correct, with a host of "ifs". I think that we need to understand all of the unknowns in order to correctly assess the situation. I’m just afraid that if the gloomy side of this is over stated, then we will have a real problem on our hands...due to panic and paranoia.
I've enjoyed your dialogs. Do you think that the pollies will join us in killing rats? Even if everyone in the world is polite to the USA come Y2K et al., the inflationary recession is coming. There will be slowdowns, blamed on Y2K even if it simply managerial incompetence. I'm saving some #11 rat shot for my bios chip.
Thanks!
On one hand, I would like to see a significant down turn, JUST so that the boomers would realize that most of them are living a lie, relying on credit to boost their life styles.
Have you ever read the millionaire next door, can't remember who wrote it. The author found that a good percentage of millionaires live very normal lives and don't buy expensive houses/cars/watches/stuff. Living well undernieth their means and hiding the fact that they have money. But todays average couple is so strung out on credit, trying to stay up with the Jones!! I think I read that savings in the US just went negative! They are going to be in a world of hurt if Y2K hits the market.
So, I am going to take a lession from that book and PRETEND that I don't have any Y2K provisions.... <;^)
Or do you mean they will all be running but with different time clocks?
That's exactly what I meant.
"I work for a major utility that supplies power and water for about a million people. We have moved all our clocks forward on all of our facilities and everything looks good although we are cutting it a little close on a certain specific control capability (but even that appears to be on track). I do not think that everything is OK everywhere, but I am tired of speculation and non-specifics and foolish conclusions regarding Y2K like this article. I would like to see specific tests and failures that certainly the programming community must be familiar with as these programs are investigated. The biggest threat to the US I see is sabatage and possible foreign supply disruptions and perhaps an economic judgement for a nation that has rejected the commandments of God."
I'll tell ya what...I'm a Conservative, Oh, I mean a Liberal, oh, what's the difference anyway? In any case, you are guilty until proven innocent. Show me the money!
"I work for a major utility that supplies power and water for about a million people. We have moved all our clocks forward on all of our facilities and everything looks good although we are cutting it a little close on a certain specific control capability (but even that appears to be on track). I do not think that everything is OK everywhere, but I am tired of speculation and non-specifics and foolish conclusions regarding Y2K like this article. I would like to see specific tests and failures that certainly the programming community must be familiar with as these programs are investigated. The biggest threat to the US I see is sabatage and possible foreign supply disruptions and perhaps an economic judgement for a nation that has rejected the commandments of God.
des = does.
Recently, we've had a major hurricane on the east coast, and a major earthquake in Taiwan. Both were genuine disasters. Both resulted in significant loss of productivity. But neither one has had a significant effect on the world economy, and neither had much (if any) effect on the global infrastructure.
Maybe not a significant effect on the world economy, but a definite significant effect on my ability to get my blown computer replaced at work. The IT guy told me they get their memory from Taiwan. The "memory" they used to buy for $69 before the natural disasters is now commanding a $500 price tag. When he confidently said, "Oh, but it will go back down & stay there once they get things straightened out over there," I almost choked on my tuna fish.
Now multiply this little scenario around the world with all our other imports when other foreign countries crash and burn. Opinion?
Now multiply this little scenario around the world with all our other imports when other foreign countries crash and burn. Opinion?
If you assume that everything is going to crash and burn, then you get a horrible situation. But everything isn't going to crash and burn. In this example, a significant portion of the capacity of Taiwan was destroyed (not just hampered with computer glitches), and the biggest problems is that you have to pay a premium for memory for a few months (I do symphonize; I was planning on putting together a couple of computers as gifts for relatives; now I have to think twice about whether I want to pay the premium or wait until next year). If it had been Y2K problems instead of an earthquake, they would have been back up and running in days, and you never would have noticed.
I wouldn’t be a bit surprised to see spot shortages of some things due to Y2K. The trouble is that no one can predict what those things will be. (I know, the doomers will say: “electricity, food, water…everything”.) The problems will primarily be in those areas that we haven’t thought of, not the ones that have been discussed ad nauseam.
At this point, practically every business system in use is already processing post-Y2K data. If Y2K were going to be the disaster the doomers are predicting, we would already be seeing significant problems. We’re not. The peak of the problems will occur during January, but you could multiply the current level of Y2K problems by 10, or even 100, and not even notice the problem. A year ago (when I didn’t have any first-hand experience to draw upon), I decided that if the problems were getting bad enough by about September, I’d go into hiding. Then I spend most of this year working on Y2K issues, and I stopped worrying about major (civilization-jarring) problems. But even if I didn’t have real-world experience, that fact that we're this far into the year without any noticeable problems would have eased my concerns considerably. Here we are, halfway through October; most major businesses and governments are already into fiscal year 2000, and the worst problem is a cosmetic typo on some vehicle registrations (see the thread about “horseless carriages”).
By all means prepare, but please don’t do things that will cause you grief if civilization doesn’t collapse (because it won’t).
John
The problems will primarily be in those areas that we haven’t thought of, not the ones that have been discussed ad nauseam.
If that's the case, I have to ask, why do you visit Y2K sites so frequently? I ran a search on your name, and with a couple of exceptions, you visit Y2K sites almost exclusively.
If that's the case, I have to ask, why do you visit Y2K sites so frequently?
I read lots of threads on FreeRepublic. I post on Y2K issues because it's the one area covered here on which I can speak with some authority. The vast majority of people (including the people here) don't have any first-hand knowledge or experience on which to draw their own conclusions about Y2K. I'm offering my experiences for people to consider when they're making their decisions about how to prepare for Y2K.
Would you prefer that I go away, so that you can wallow in fear or panic?
Keep well.
John
DouglasKC,
Thanks for your comments. Although I disagree with some of your conclusions, I appreciate your thoughtful response. I think that way too much bandwidth has been wasted with 3rd- and 4th-hand information, and insults and name calling. It's nice to get some actual first-hand observations posted here.
Keep well.
John
Nothing significant happened on 1/1/99, 4/9/99, 9/9/99***
First, what do these dates have to do with Y2K and second, 9/9/99 did cause problems and, yes, they were severe. One of our police departments went down (in the Detroit area) and two of our District Courts had all of their files deleted. Just because incidents aren't reported at length in the media, doesn't mean nothing happened.
First, what do these dates have to do with Y2K and second, 9/9/99 did cause problems and, yes, they were severe.
1/1/99 was significant because some programs were written to use a year value of "99" (regardless of the month or day) to indicate error or invalid conditions (this one caused a couple of minor problems where I worked at the time). 4/9/99 might be stored as "DDDYY", which would be "09999" and some theorized that the four 9's could cause problems (I'm doubtful).
It's interesting that you say that "9/9/99" caused problems; even most of the doomers here discounted that date as a problem.
One of our police departments went down (in the Detroit area) and two of our District Courts had all of their files deleted.
How certain are you that these problems were caused by Y2K bugs? Systems crash every day. Just because they crashed on 9/9/99 doesn't necessarily mean that they had Y2K problems. (I've seen both occurrences on numerous occasions, but since there wasn't any Y2K hysteria at the time, it never occurred to me to blame it on that.)
By the way, what's your definition of "severe"? Are they still down? Did people die, or suffer serious problems, because of it? I have two relatives who have worked as police dispatchers; and they've told me of times when their computer equipment failed (nothing to do with Y2K). It made their jobs harder than normal, but they managed just fine.
I didn't pay much attention to the 9/9/99 hype. I'm not a computer expert but people who were said there shouldn't be any problems and this didn't have much to do with the Y2K. Frankly, I was more concerned about the utility (electric) test that day.
The people who worked on the computers said the problems were date-related. I don't know the status of the courts, but the police dept was fine by about noon.
First, people dying should not be the criteria for "severe". I think inability to function is severe. In the case of law enforcement, the more knowledge you have of an individual, the better able you are to make decisions regarding handling of the individual. Sure police didn't always have computers, but more police officers were shot or wounded.
First, people dying should not be the criteria for "severe". I think inability to function is severe.
Given the number of computers used by the police throughout the country, it's likely that some of them are unable to function at any particular time. Although that's unfortunate (and may at times put officers at additional risk), it's also a normal part of life.
Note what didn't happen: the system didn't stay down for an extended period (I'll bet it wasn't the first time that their systems was down for at least half a day), and it didn't bring about a systemic failure of other police departments.
My guess is that when the computers went down, the reactions of the users wasn't, "Oh my God! We're doomed!" The much more likely reaction was, "Stupid computer is down again; where did we put big maps and notepads from the last time this happened?"
You said that you weren't a computer expert. I suspect that means you don't realize how common it is to have computer problems of all types. Fixing problems in existing computer programs is the normal activity of most programmers; developing new stuff is the exception. Of course there will be problems associated with Y2K; but we've had a lot of experience fixing such problems.
Keep well.
John
I am not a computer expert, but the following story impressed me with the likelihood of many rather than few failures in things that will affect our lives:
About two months ago, a suburb of Los Angeles decided to test its waste processing plant for Y2K compliance after its systems had been made 'compliant'. They advanced the clock. The plant promptly dumped several million gallons of raw sewage onto the city park's grounds.
"One big difference is that in a strike, people are purposely trying to disrupt production. In any Y2K problem, they will be actively working to minimize any disruptions. Also, while that may have been painful for the people involved, the vast majority of people never even noticed that strike."
I don't think your big difference is all that big. After all, management and union were working to resolve the differences and restore production just as programmers will be working to resolve any Y2K bugs. While you are correct in pointing out that few noticed the effects of the strike this doesn't disprove the analogy either. If Y2K effects occur, they are are unlikely to be limited to a single industry or region or even to the private sector.
I'm reminded of another labor dispute of the late '70s that had major effects here in North Alabama involving a truckers strike. The truckers parked their rigs around the only gasoline distribution point blocking access to the facilities that filled tanker trucks. Gasoline lines like the Arab oil embargo quickly formed. The authorities removed the trucks and distribution was restored. The effects lasted a couple of weeks. The effects would be the same reguardless of the underlying cause.
It seems to me that you're argueing that contingency plans can mitigate the effects of Y2K reguardless of the extent. I don't doubt that this is true as long as the mitigators are not swamped. Suppose the plan calls for a Purchasing department to rely upon manual workarounds requiring additional personnel. No problem if it is only purchasing. But if multiple departments require additional personnel to implement their contingency plans, where are all these personnle to come from?
I heard about this story. The kicker is that the sewage was dumped right where the town is having its New Year's Eve party! :-)
Obviously, if they had not properly tested their Y2K bug "fixes" then they weren't Y2K compliant (the Y2K compliance standards I've seen require successful final testing to achieve compliance, not the word of some programmer that it's compliant). But just as obviously, they did test their work and found the problems two months ago, rather than on 1/1/2000. At that point, I'm certain that they made the programmers go back and do their jobs right, and retested to make sure it really worked.
So, while this must have been a huge embarrassment for the programmers and managers involved, the Y2K testing worked; it did exactly what it was supposed to do.
However, let's assume that they hadn't properly tested their system. The net result would have been a smelly embarrassment on 1/1/2000, but it wouldn't have been a crisis.
Keep in mind that this suburb wasn't the only city to test their sewer systems for Y2K compliance, but they apparently were the only one to find a problem serious enough to actually leak sewage.
It's very common for a bug fix in a computer program to break something else in that program; computer programs are very intricate things. Which is why you test them after you think you've fixed them (which this city properly did).
Keep well.
John
However, let's assume that they hadn't properly tested their system. The net result would have been a smelly embarrassment on 1/1/2000, but it wouldn't have been a crisis.
I'm glad they tested them early too. Some places are not. It's more than a smelly mess, and it can cause sewage to back up into houses. If you have raw crap in your bathtub, that's not just embarrassing; it's life threatening. Seriously.
The bad news from Van Nuys is that other systems may be hesitant to test because of the bad publicity. This is the biggest danger of Y2k; looking good as long as you can.
It's more than a smelly mess, and it can cause sewage to back up into houses. If you have raw crap in your bathtub, that's not just embarrassing; it's life threatening. Seriously.
I suppose that, in theory, it can be life threatening, but sewer backup aren't that uncommon (happened to my parents recently, in fact), and I've never heard of someone having their life threatened by it (I'm not saying it doesn't happen, but it's not the normal result of having a sewer backup).
The bad news from Van Nuys is that other systems may be hesitant to test because of the bad publicity.
I've never met a manager who would react even remotely like that (but I've never worked for the government!). The biggest problem I've had with bosses is that they overreact, rather than underreact, to potential Y2K problems. (And I better explain that, or the doomers will jump down my throat. To overreact is to disregard the seriousness or the likelyhood of failure when allocating resourses. Example: one of the managers at my current job wanted to stockpile hard drives, in case our current hard drives weren't Y2K compliant. Anyone who knows anything about computer hardware knows that hard drives don't know or care what the date is, and are therefore immune from Y2K problems, even if the computer they're attached to aren't.)
Keep well.
John
"About two months ago, a suburb of Los Angeles decided to test its waste processing plant for Y2K compliance after its systems had been made 'compliant'. They advanced the clock. The plant promptly dumped several million gallons of raw sewage onto the city park's grounds."
Not quite true. The test was to see what would happen if there were a power failure; a power failure due to earth quake, over load, Y2K, or a host of other reasons. And yes, the power failure caused a dump of raw sewage. This is a failure on the part of whomever designed the sewage system and is only related to Y2K because Y2K has a potential to disrupt power. A very loose linkage, if you ask me!
I was in L.A. when the event happened and that's the way it was reported on TV at the time. There could have been more to it, however, as you imply.
I hope you took no offense.....it wouldn't surprise me if the story was inaccurately reported. Now, which one of us got the wrong story? knowing how well the media report stories, it could be either, or BOTH of us!
Anyone else out there hear a good version of the "Crap in Van Nuys" story? ;)
Ok, just one bump from my Y2K gloat folder. Again, the doom here was being passed by an entity that was selling a product. A rather vile form of advertising, if you ask me.
How cute. You saved all the Y2k threads for a Bump-&-Gloat orgy.
how many companies are in a rush to advertise that they are having Y@K problems?
How cute. You saved all the Y2k threads for a Bump-&-Gloat orgy.
Hey, we're not all gloating.
If somebody hadn't held up the scary prospect of people freezing during power shutdowns
in Minnesota or riots in Los Angeles, nothing would have been done to get the computers
fixed up for the date change.
It's just like the stock market, whenever people are saying it's in trouble, I usually
buy more stock.
When people are really optomistic and think the ecoonomy will expand forever...
then I get scared.
Ok, just one bump from my Y2K gloat folder. Again, the doom here was being passed by an entity that was selling a product. A rather vile form of advertising, if you ask me.
How clever of you to save a Y2K gloat folder.
Having lived in earthquake country in California for many years, my family laid-in a fairly good supply of extra food, water and bought a generator. When we left CA, we gifted our supplies to some friends who found them invaluable during the Northridge earthquake.
As for the generator, even though we have always lived in major metropolitan areas, in the last ten years, it has been used at least ten times when the power was off anywhere from 30 minutes to a full day. Amortized, the generator has been a great investment. We are once again stocked up with extra food and water. Now living in tornado country, that may too come in handy.
You can mock Y2K preparedness/preparers all you want, sonny. But those of us who have "been there, done that", know who is really the fool.
I did not mock preparedness. This was a comment on those that "trumped" up the probability of a disaster. The assertions in this article are not true. Prepare all you want, its a hell of a good idea. I'd just hope that folks could make preparation decisions base on good information.
Scaring people to sell a product, however, is damn near criminal.
Ok, just one bump from my Y2K gloat folder. Again, the doom here was being passed by an entity that was selling a product. A rather vile form of advertising, if you ask me.
That is rather cute. Your own Y2K gloat folder. Good for you. Do you really think that if the worst case scenario had happened that any gloomer (well, maybe Jethro :- )) would come on here and "gloat"?
Here's an analogy for you.
Three guys Bob, Ray and Fred are discussing cars and driving. Ray has never driven a car before but wants to.
Fred urges Ray to prepare. He tells Ray that he should take driving lessons, get insurance, buy a safe car, wear his seat belts, obey traffic laws, etc. etc. etc., because it's potentially dangerous to operate a vehicle. In the worst case, the car might crash and he could die.
Bob thinks that preparation like that is overkill. After all, he's driven before and never had an accident. The other drivers will watch out for him. He can learn on his own and he probably won't get pulled over so he doesn't really need a license. Besides, he hints, Freds brother is a driving instructor...Fred just wants to cash in.
Ray listens to both sides, and decides to take Freds advice and prepare. Finally, the big day comes. Ray goes out on the highway and drives around. He doesn't drive perfect, there are a few glitches, but he makes it and doesn't crash and is rightly quite proud of himself.
Fred of course is glad that everything went okay and feels good that he urged Ray to prepare for the big event.
But strangely, Bob proceeds to gloat. "I told you that you wouldn't crash!" he screams with glee, "I was right! You must feel like a real idiot Ray falling for Freds doom and gloom scenario. Look at all that time and money you wasted on unneeded preparation."
Can you figure out who you are Tanker?
You can mock Y2K preparedness/preparers all you want, sonny.
I don't know of anyone who "mocked Y2K preparedness/preparers". At least I've never seen any, and certainly haven't seen any here. For that matter, I doubt that there's a single person here who didn't do *some* level of preparation themselves.
Saying that people are/have been nastily mocking Y2K preparation, and especially the implied claim that that has been the primary focus/reason for the "gloat" threads, is a straw man.
What *has* caused people to enjoy gloating was the fanatical extremism of some of the Y2K doomers. It's really hard to hold back the "I told you so's" when some (not all) of the doomers were, until today:
1. Ridiculing perfectly rational analyses by those of us who were "in the trenches" in the Y2K bug-fixing field, pointing out why minor disruptions were to be expected, but nothing that would kill thousands of people, or end civilization as we know it, or justify going completely overboard over the thing. I, for one, got tired of being called a shill for the government just because I pointed out that most of the bugs have already been fixed, and most of the rest wouldn't disrupt anything.There's nothing wrong with prudent preparation. But the more extreme doomers have (prior to today) edged into the territory of being real assholes about the whole discussion, and I for one am enjoying the opportunity to go "neener neener, you were wrong and we were right. Very seldom in any contentious discussion is one side vindicated so clearly and suddenly -- we might as well enjoy it for a few days. Most other types of arguments never get to be settled one way or another.2. Lots of doomers were crying, "you'll see!!" That's a challenge that we're quite happy to settle. After years of being arrogantly told that we'd come crawling to them in desperation after The Big One, and being told that *then* we'd see the light, it's only human nature to now ask *them* if *they've* finally seen the light we've been trying to shine on the topic for so long. Both sides long ago agreed to compare notes, and declare a winner, once the date rolled over.
3. Some of the doomers went so far as to say (explicitly or implicitly) that they'd enjoy watching us starve to death when we were proven "wrong", and/or shooting us for trying to reach some of their own supplies. I think a little bit of "I told you so" in return is *hardly* even in the same ballpark as what we've had to put up with ourselves.
You saved me from needing to point out what you did on post 34. If those "mission" critical systems share common methods of doing things then those systems become parallel in nature. Either they all fail from the same cause or none of them fail. Calculating the actual probabilities of failure is much more complicated. The writer's assumption that each system is completely independent and different in structure is flawed. Within the same company ideas/structures are often replicated to save the cost of reengineering each system that has functions in common. The point of this is that if you don't include the commonality of interdependent systems, you over estimate the probability of failure.
I design hardware and software mostly used in "embedded" systems. In all of the applications I've done, the current time was only used for maintenance purposes. Basically logging events in a large communications network. A clock failure would not result in any loss of data traffic in these systems. It would only hamper in trouble shooting a failure, especially an intermittent failure, by making it more difficult to locate a faulty piece of equipment.
But strangely, Bob proceeds to gloat. "I told you that you wouldn't crash!" he screams with glee, "I was right! You must feel like a real idiot Ray falling for Freds doom and gloom scenario. Look at all that time and money you wasted on unneeded preparation."
Excuse me, but this is a hugely inappropriate analogy. I think you owe Tanker a big apology.
If you think that the doomers were saying nothing more unreasonable than, "you should learn to drive, and have insurance, before you drive a car", then you are *way* out of touch with the actual dynamics of the recent hysteria.
A far better analogy would be the recent nuclear-powered Casini spacecraft launch, and the hysteria that produced among the more nutty members of the environmentalist movement:
Doomers: "It'll crash into the Earth! Plutonium will be released!! We'll all die of cancer and radiation!!!"Pollys: "Look, we've taken every precaution -- we've built it as best we can, even if the launch goes awry the containment vessel will stay intact, and as far as the Earth flyby goes, a navigation failure would be a million times more likely to make it miss the Earth than to hit it. And even if the thing did crash and break open, it would be a tiny spot of radiation on a very large Earth. Relax, we'll likely come through it just fine."
Casini then launched, and did its Earth flyby, flawlessly.
Doomers: "How ***DARE*** you take such an enormous chance with the lives of everything on Earth!!!"
Pollys: *Sigh*.
Do you really think that if the worst case scenario had happened that any gloomer (well, maybe Jethro :- )) would come on here and "gloat"?
Yes, I most certainly do.
A large number of them were gloating *before* the fact, predicting their moment of vindication, and our comeuppance, when things finally went to hell.
That explains why a lot of us now enjoy rubbing their noises in it.
I repeat -- no one that I know of is giving a hard time to people who simply thought it prudent to prepare. I prepared myself. The people we *are* thumbing our noses at are the ones who have been making jerks of themselves for a year or two now, and who now deserve to have their noses rubbed in it for a day or two as a small amount of payback. *And* as a not-so-gentle reminder that perhaps they might not want to go so far overboard the next time they get all excited about the Next Big Coming Disaster, whatever it may be.
If that makes you feel defensive, then I have to wonder why...
Well said. Doing an actual failure analysis (that is accurate) is usually much more difficult than the design of the actual system being analyzed...
People go on here about that "small business" that didn't prepare for Y2k and how it could bring the big car manufactures (for example) to a dead stop for days… I own half a small "high tech" business and if one the computers used throughout our business craps out because of a hardware/BIOS Y2k problem we'll simply throw the computer out and get another… Yes it would be a bit of a pain in the rear, but no big deal. Computers die all the time. They are imperfect machines. Anyone that is completely out of business because their computer died has more problems than Y2k. There is a big difference between problems and catastrophic failure. Thanks for sharing your knowledge.
Test yourself! If Company A has 200 mission-critical systems, and each mission-critical system has a 3% chance of failing on January 1, what is the overall chance of the entire company failing on January 1?
I got the math right, but your assumptions are dead wrong.
First, very few companies have a large number of "mission-critical systems". Most things can slide, if they really need to.
Second, very, very, very few things (far fewer than the "3%" you posit) can truly "fail" on January 1st. Most things would just be hung up for a bit, and then fixed as people jump on the problem and take care of it one way or another.
Finally, most companies are very, very aware of what their "mission critical systems" are, and long ago spent whatever resources were necessary in order to make bloody sure that they *were* Y2K ready. I know *I* did.
Here's a better restating of your "pop quiz" question, which more accurately reflects reality:
If Company A has 2 mission-critical systems, and each mission-critical system has a 1% chance of being at all dependent on the current year, and a 1% chance of having been overlooked during Y2K testing, and a 0.1% chance of actually failing on January 1, and a 0.1% chance of not being able to fix the problem once it manifests itself before things fall apart completely, what is the overall chance of the entire company failing on January 1?Answer: Pretty f*cking unlikely.
Reality check: Know of any companies that entirely failed on January 1? Neither do I.
*This* sort of know-nothing nonsense is what caused a lot of us to get really annoyed at the constant nattering from idiotic doomers a long, long time ago.
Yes, I most certainly do. A large number of them were gloating *before* the fact, predicting their moment of vindication, and our comeuppance, when things finally went to hell.That explains why a lot of us now enjoy rubbing their noises in it.
Who are they? The only one who I can recall being blatent about it is Jethro. Everyone other "gloomer" on Freerepublic seemed to be levelheaded and matter-of-fact.
So just who are you talking about? If you care to read this entire thread, you can see that many "doomers", myself included, are quite reasonable in our posts.
So let us know please.
Reality check: Know of any companies that entirely failed on January 1? Neither do I.
Snort...do you know any companies that were actually operating at the rollover? Hard to fail if you're not open. And it's not likely they're going to have any press releases stating that their computer systems are screwed up due to Y2K.
*This* sort of know-nothing nonsense is what caused a lot of us to get really annoyed at the constant nattering from idiotic doomers a long, long time ago.
There ya go, calling names.
You spoke my sentiments exactly. Thank you. A few here earned what they have been receiving. Payback is a bitch.
It amazes me how many here apparently can't comprehend what they read. Over and over we have stated that we are directing our comments at those who called us every name in the book over the last year or so for not agreeing with their assessment of Y2k. Remember the "white busses" and FEMA… The "billions and billions" of embedded chips that were going to fail and knock out the telephone and electricity… How it was going to cascade across the country… That these chips could not be replace in time because greedy business was ignoring the problem and that testing them was "impossible". I don't know of a single post that suggested one should not prepare at all or that one was a fool for doing so. It is the "bunker" people that assaulted us for trying to provide information to the contrary that we say, "TOLD YOU SO".
Good for you.
Spectator bump.
Snort...do you know any companies that were actually operating at the rollover
Yes.
All the electric utilities in the world.
All the phone companies in the world.
All the cable/satellite TV providers in the world.
All the cable channels in the world.
All the internet providers in the world.
All the gas companies in the world.
Endless Fast Food restaurants in the world.
Nearly every multinational company has on going business/data processing 24 hours a day 7 days a week.
This list could go on and on but why bother… You're trying to prove a point that the evidence does not support. You have a tough row to hoe.
The premise of his article is incorrect. Read on if you wish to know why.
So you're telling us that as you were fixing the credit card problems of this company you were breaking lending/credit card consumer laws.
Is that what you're telling us?
It's for this very reason the strictness of these consumer lending laws that corporate lawyers have told the techies not to admit to any problems with y2k.
P.S. Every state has their own auditors audit all consumer credit companies that have a business presence in their state. It is most difficult to deal with them under normal business activities but to have y2k problems affecting the credit products that consumers use could be crushing.
So the word is on the street.
It amazes me how many here apparently can't comprehend what they read. Over and over we have stated that we are directing our comments at those who called us every name in the book over the last year or so for not agreeing with their assessment of Y2k
Then kindly do the research and address the posts to the people that called you names. This was bumped with a reply to me, so I am going to respond.
Actually, the post you quote part of was to Dan Day, reply #79 who was responding to Jolly Green not you. You haven't called me any names that I know of so far. I disagree with the premise of your article and I said why much further up. I wasn't being rude to you. Just pointed out why I disagree. TankerKC did a much better job of describing why he disagreed than I but we had the same reasons.
I'm sorry about the "To #" in the original post you quoted. I used the wrong "Reply" button to reply to... I did have the right name though... I hope that made sense...
I know who I am without your inappropriate analogy, thanks.
This was never intended to be a "Gloat at Doug" thread. I wanted to highlight what I believe was a error in the original article. An error that was not put in by accident, but by design. The author(s) overstated the probability of failures in order to sell their Y2K survival kit. It had nothing to do with you. In fact, I think we agreed to disagree earlier in the thread.
All we could do be before was wait and see. Well, now we waited. See?
Most Respectful Regards
Another interesting point:
I went back to review the article on the Y2k Newswire--it is gone.
So, I thought I'd review the probability of Failure Chart--it was gone.
But, what of Arial Marketing, Inc. (the "authors" of this ground breaking piece of science/advertising)?
A web search yields very few links that are still active. (Imagine that!). One that is still open is here.
One quick quote: ” Our numbers predict an 87% chance of a collapse of modern civilization."
Midday bump with a question. Curiosity made me look for the original article, but it is gone. If fact many “gloomy” articles are gone. Are any of you familiar with Arial Marketing or the folks who run the Y2Knewswire.com site?
Exactly, this guys assumptions are flawed, what is a Mission Critical System? Well many companies think EMAIL is a mission Critical system, and it is important, but if EMAIL is down for 5 hours because of a Y2k "bug" Does this mean the entire company fails? Of course not. I have no doubt there will be small problems related to this, but the doom and gloom predictions are pretty much bunk.
Does that mean that there will be no problems, NO. But if you think one system, or using this guys math 6 out of 200 go down that the entire world is going to collapse not bloody likely.... fact is systems go down all the damned time, I KNOW, i am responsible for keeping them running, and things continue.
I would suggest watching your financial statements more closely then you may do otherwise etc, but it is not the end of the world as we know it.
Reality check: Know of any companies that entirely failed on January 1? Neither do I.
Snort...do you know any companies that were actually operating at the rollover? Hard to fail if you're not open.
Oh, *NO* you don't... I am *not* going to let you get away with that one...
Look, guy, *YOU* were the one who posted the article that brought up the issue of, and I quote, "the overall chance of the entire company failing on January 1".
Not January 2. Not January 3. Not sometime this year due to snowballing problems. "January 1".
Now, when I respond exactly in kind, and point out the lack of "entire companies failing on January 1", *NOW* you suddenly want to shift the discussion and mumble about companies allegedly not "actually operating at rollover" (despite the fact that many of them actually *were*).
Don't you even read the articles that you yourself post??
Don't blame *ME* for sticking to the same topic that YOU YOURSELF introduced here.
Sheesh.
And while we're at it, I notice a distinct lack of "entire companies failing" even now that everyone's back to work.
*This* sort of know-nothing nonsense is what caused a lot of us to get really annoyed at the constant nattering from idiotic doomers a long, long time ago.
There ya go, calling names.
When the shoe fits, you bet.
The original article you posted was remarkably arrogant and condenscending, claiming that most people (especially the experts looking into the Y2K problem) were incapable of grasping the eight-grade math that the author presented as some sort of Deep Analysis, while at the same time being laughably unrealistic in the rest of his depiction of how companies do or do not work.
So yeah -- juvenile analysis on the 13-year-old level, coupled with an attitude that the author has a better grasp of things than the real specialists on the subject, deserves all the abuse it gets.
When I called it "know-nothing nonsense" and "constant nattering from idiotic doomers", I was being kind. Would you like to hear what I *really* think?
Exactly, this guys assumptions are flawed, what is a Mission Critical System? Well many companies think EMAIL is a mission Critical system, and it is important, but if EMAIL is down for 5 hours because of a Y2k "bug" Does this mean the entire company fails? Of course not. I have no doubt there will be small problems related to this, but the doom and gloom predictions are pretty much bunk.
Exactly.
When things failed to fall apart on January 1, a lot of the doomers started saying, "oh yeah, well just wait until everyone goes back to work on Monday!"
My standard response is, "any company that can afford to shut down for the weekend without collapsing can afford to have Y2K glitches hold them up for another day or two without disaster as well".
Only things like utilities or nuclear weapons systems could have caused havoc with even an hour-long glitch. Any other company can afford to declare a long weekend, if necessary to fix bugs, without fear of crashing and burning.
Companies routinely have to miss days because of weather (snowstorms, hurricanes) or other unforeseen circumstances, and if *that* doesn't force them into bankruptcy, then neither will any previously undiscovered Y2K booboos.
I'm baffled and have egg on my face. I was truly concerned, and urged everyone else to be also. It seems to have been over nothing. Even the "glitches" (I hate that word) don't seem extraordinary at all.
I just need an explanation why some countries spent billions and others spent almost nothing with same results. The "bug" wasn't fixed, at least on a global scale. Yet it caused some small problems. Did the big corporations spend hundreds of millions for nothing? Testing showed failures. That is a fact. But it didn't happen in real life, except for a little bit. Just enough to make me not dismiss it. I'm still sitting here thinking "WHAT THE HELL JUST DIDN'T HAPPEN?"
I've seen a real lack of squirrel-tossing right-wing terrorists and 10 zillion (for you, Nita) email viruses.
Admitting I was wrong, but sitting here puzzled...
Some more math:
Assume a company has one "mission critical system", with a 3% chance of Y2K rollover failure.
Probability that their ops will stay "up" = 0.97
Probability that their ops will go "down" = 0.03
Now assume two "mission critical systems", each with a 3% chance of Y2K rollover failure.
Probability that system "A" will go down = 0.03
Probability that system "B" will go down = 0.03
Probability that ops will completely go down = 0.0009 (.03 x .03)
Probability that ops will at least partially stay "up" = 0.9991
Probability that ops will stay completely "up" = 0.94
Now assume three systems:
Probability that each system will go down by itself = 0.03
Probability that ops will completely go down (all three fail) = 0.000027
Probability that ops will at least partially stay "up" = 0.999973
Probabilithy that ops will stay completely up = 0.912673
You can follow this on through, but the conclusions are beginning to become obvious:
The original article behind this thread was only partially right: as the number of systems at some risk of failure in an enterprise increase, then the probability that the enterprise will escape with NO systems failures whatsoever decreases.
That they failed to either understand or to tell you, however, is that as the number of systems increases, the probability of ALL of them failing DECREASES.
In fact, this little exercise illustrates why complexity can quite often contribute to GREATER stability and "robustness" rather than increased vulnerability. It is the same reason why the ecology of a rain forest is more stable than is an arttic island populated by lemmings. And it may be THE reason why we did not see cascading failures taking everything down at the rollover.
Nevertheless, given some level of probability of systems failures, however small, and given the population of systems out there, it is inevitable that we will see some numbers of system failures. And we are.
And it would have been a whole lot more helpful if EVERYONE had understood that basic math all along!
I just need an explanation why some countries spent billions and others spent almost nothing with same results. The "bug" wasn't fixed, at least on a global scale. Yet it caused some small problems. Did the big corporations spend hundreds of millions for nothing? Testing showed failures. That is a fact. But it didn't happen in real life, except for a little bit. Just enough to make me not dismiss it.
A number of factors were at work:
1. I think the "amost nothing" descriptions of other countries' Y2K preparations is understated.2. Countries by and large spent proportional to their level of computer usage. Countries that are heavily dependent on computers, like the US, spent a lot. Countries that weren't, didn't have to spend as much to get similar results.
3. The US bore a larger than average share of the burden, because a lot of places buy computers and software from *US*. We had to spend money to fix our products. The buyers only had to receive fixed versions, but didn't have to shell out dollars for bug-finding and bug-fixing.
4. You can pay now, or you can pay later. The US, and some other countries, chose to spend the money up front in order to head problems off at the pass. Some other places chose to just see what broke, and then pay to fix it after the fact. This would have been a bad strategy if most Y2K bugs were likely to be of a disastrous nature, but they aren't likely to be. Consequently, the "nothing exploded" news reports make it look as if the procrastinating nations got off scott free, but they didn't -- they'll still have to fix a lot of niggling problems that the US for the most part avoided having in the first place, due to planning (and spending) ahead.
It is the same reason why the ecology of a rain forest is more stable than is an arctic island populated by lemmings.
I don't think you meant that to be funny, be it's apropos. LOL.
I agree with everything you said. Except for the embedded chip question. There should have been failures, bigtime. At least someplace. Everyone worried about it, but the final consensus was that it would affect very few.
Wrongo. None. .0001% still would have been huge.
Don't get me wrong. I am absolutely thrilled and have a tremendous weight off my shoulders. Except I feel stupid.
To me, this is the best reason not to believe that Y2K is going to be just a "bump in the road."
Douglas, Y2K has ALREADY shown to be much more than a "bump in the road." But your analysis in computing the probabilities of a failure neglects an unrelated (but important) factor: failures, when they occur, will simply NOT be reported or else be attributed to viruses, hackers, terrorists, etc.
...10 zillion (for you, Nita) email viruses.
Forget the 10 zillion viruses, worms and trojan horses--I don't want 'em. I'm still waiting for 10 zillion dollars to drop in my lap from somewhere.
Ahhhh..... I can dream, can't I?
One must not forget that there were tens of thousands of humans standing by to bypass glitches because glitches of this type were expected. We will never know how many systems stuttered, but were manually over-ridden or skirted contemporaneously or prior to the main event. Any dooming to excess may have awakened a few to be ready for the next comet impact. {8^)
Life is like a box of chocolates. It is nice to have a box or two around sometimes.
I work for a major utility ...
OK. Granted that your utility is 100% compliant; what will you do if the trains can't deliver fuel?
Now grant that the railroads are 100% compliant; what if the mines in Indonesia can't produce coal?
Now grant that the mines are Y2K compliant; what if the the transport ships' cannot navigate due to Y2K?
See where this is leading? Your major utility is NOT the only entity necessary to keep the lights on. The only thing you know for sure is: No fuel - No electricity.
I agree with everything you said. Except for the embedded chip question. There should have been failures, bigtime. At least someplace. Everyone worried about it, but the final consensus was that it would affect very few.
The whole "embedded chip" issue was overblown from the start.
Yes, many, many types of electronics these days have embedded chips. No, most of them don't give a rat's rear end about the date.
A fancy toaster may have an embedded microchip controlling the toasting cycle, but it doesn't know (or care) what the date is. Thus, 01/01/00 is just the same as 12/31/99 to it (or for that matter, 06/08/52). Embedded chips of this sort are 100% impervious to any Y2K problems.
Someone on the DVD discussion group heard a rumor about certain types of DVD movies being non-Y2K compliant, but again, DVD players don't know, or care, what the date is, and thus it's impossible for them to fail just because the date happens to roll over.
And the same goes for most other embedded chips and microprocessors.
The only embedded chips that could fail because of Y2K are the ones that know, and care, what the date is. And that's a very tiny percentage of embedded chips. And the very fact that they *do* need to know what the date is makes them obvious -- there has to be a way to *tell* it what the date is. A good example of this sort of embedded chip is the guts of your VCR. It has to know the current date so that it'll know whether you wanted to tape that show today, or next Tuesday. And because it cares about the date, your VCR has a screen where you can set (and view) the current date.
But this same necessary feature makes it easy to detect, and work around, any potential Y2K problem. If it malfunctions, just reset the date manually. If it can't take a 2000 date at all (rare), punch in 1972 instead, which had the same day layout as 2000. If *that* fails, enter 1995 (IIRC) until leap day (Feb 29), then switch to 1996.
Furthermore, most embedded chips are Y2K compliant, because people who build hardware of that sort know very well that they can't update the things later (like is possible for software), so they're more likely to plan ahead for things like Y2K. Plus, embedded microchips are a much more recent development than software programs, so most of the embedded chips out there were made close enough to Y2K for Y2K-compliance to have been on the developers' minds (unlike some of the relatively more ancient software programs, which can date back up to 30+ years ago).
Finally, a lot of embedded chips care about the time of day (e.g. my car's digital clock) or the day of the week (e.g. my electronic thermostat), but don't care about the actual *date*. The number of embedded chips which actually care about the year is quite small. The number of them which can't be reset to 2000, or an equivalent year, is *very* small.
The whole "embedded chip" issue was overblown from the start.
I don't think it was. Are you familiar with the IEEE organization? If not, they are a society of electrical, electronic and computer professionals that develop standards for these industries. I will give you a number of links. After reading them, tell me how they, a group made of of people that design, build, and develop test procedures for embedded systems, could have been so wrong.
This first link is their position paper on Y2K. It gets into the time frame of when to expect failures and the natures of failures.
This link gets indepth on embedded systems and the Y2K problem.
Here in describing embedded systems, the IEEE calculates the risk of microproccesor controlled systems (date sensitive, stand alone) failing. They said "It seems that on average about one in 100 of such systems may fail in a way which would have serious business impact. There are more faults but the resulting failures are considered to be “cosmetic” .
Later in the same link, they have this to say:
c) large scale embedded systems in these, a computer is used, more or less obviously, to control the operation and inter-working of a number of different devices used in a single process (or multiple related processes). The defining feature is that they use a hard disc to store information. Because these systems are more complex and involve many different components (any one of which may cause the entire system to fail), they are at greatest risk. Failures are most likely to be found in the application-specific software. About 20-30% of such systems have a fault: one in three of the faults would cause a significant failure (so that about one in ten of such systems poses a real risk to its owner).
There is a ton of information here, which would seem to point toward major difficulties for business. This is where I based most of my conclusions about Y2K. These people aren't alarmists, or yo-yos, or fearmongers, or hucksters. Something weird is going on. I really am curious how you can answer this.
Look, guy, *YOU* were the one who posted the article that brought up the issue of, and I quote, "the overall chance of the entire company failing on January 1". Not January 2. Not January 3. Not sometime this year due to snowballing problems. "January 1".
Whatever. If you don't want to read the entire thread to get my stand, fine. If you want to make it a prequisite that anyone who posts an article has to agree 100% with everything in it, then FreeRepublic is going to be devoid of posts.
Suffice to say that I base my opinions on more than 1 source.
So there.
Look, guy, *YOU* were the one who posted the article that brought up the issue of, and I quote, "the overall chance of the entire company failing on January 1". Not January 2. Not January 3. Not sometime this year due to snowballing problems. "January 1".
Whatever. If you don't want to read the entire thread to get my stand, fine. If you want to make it a prequisite that anyone who posts an article has to agree 100% with everything in it, then FreeRepublic is going to be devoid of posts.
You're getting more and more bizarre here.
You posted an article, I responded TO A CLAIM MADE IN THE ARTICLE.
Rather than just let it lie, on the grounds that my complaint was with the article's author, instead of the person who posted it (you), you chose to get defensive, and tried to make it seem that I was being unfair for sticking to the very same point the article itself had made.
I came back and pointed out that if you're going to post an article, you should be prepared for criticisms of it, *and* should be familiar enough with your own article to recognize when I was addressing a point made IN the article.
NOW, you're getting even more defensive, on even more tangential points, and whining about how you think I'm apparently saying that you shouldn't post articles you don't agree with.
For the record:
1. Post any damned thing you want.Suffice to say that I base my opinions on more than 1 source. So there.2. Learn how to accept criticisms of those articles without taking it personally.
3. Understand that I never said that you should not have posted it.
4. Understand that I never said that you had to agree with an article you post, *OR* that I even thought that you necessarily agreed with it (up until the point you got neurotically defensive about any criticisms of it, in which case it certainly *does* look as if you have a personal stake in the point being dismantled).
And suffice to say that I wasn't *talking* about your opinions, I was pointing out the errors in THE ARTICLE, and you need to stop being so strangely offended about that. "So there" yourself.
But hey, if you want to talk about your opinions, instead of the article itself (and your getting upset over criticisms of it), we can do that too. Earlier, you stated: "Here's what I expect: A 6 to 12 month of real economic hardship, precipitated by Y2K failures." Do you still stand by that prediction? Yes or no.
We could also discuss your wild overstating of the IEEE estimates, mischaracterizing them as "when a fireman tells you your house is going to burn down..." Shall we concentrate on that one? I challenge you to find anything in the IEEE Y2K pages that you yourself referenced, which concludes anything even remotely as drastic as "your house is going to burn down". I'll wait.
Oh, and to answer the question that you've asked others twice in this thread, yes, I do know what the IEEE is, so I'll save you the trouble of asking for the third time.
The whole "embedded chip" issue was overblown from the start.
I don't think it was. Are you familiar with the IEEE organization?
Don't presume teach your granny to suck eggs, son. Yes, I'm quite familiar with the IEEE.
If not, they are a society of electrical, electronic and computer professionals that develop standards for these industries. I will give you a number of links. After reading them, tell me how they, a group made of of people that design, build, and develop test procedures for embedded systems, could have been so wrong.
They weren't wrong. Your misunderstandings of what they wrote are what's wrong. Let's look them over:
This first link is their position paper on Y2K. It gets into the time frame of when to expect failures and the natures of failures.
Summary: "The timeframe is around the 2000 turnover, and we need to fix things before then." Hardly a portent of doom, it seems.
This link gets indepth on embedded systems and the Y2K problem.
...most of which is *extremely* general, describing in vague terms how one goes about doing a risk-assessment analysis, and builds a team to go about fixing whatever problems may be found. Again, hardly the stuff of "a fireman telling you your house will burn down".
Here in describing embedded systems, the IEEE calculates the risk of microproccesor controlled systems (date sensitive, stand alone) failing. They said "It seems that on average about one in 100 of such systems may fail in a way which would have serious business impact. There are more faults but the resulting failures are considered to be “cosmetic” .
Ooh. Aah. 1% of embedded chips WHICH ARE KNOWN TO BE DATE-SENSITIVE (an extremely small subset of all embedded chips) "may" fail in a way that would have a "serious" business impact. For example, the email system might go down. Boy, howdy, I guess it's time to run for the bunkers. So when does the "house burn down"?
Later in the same link, they have this to say:
c) large scale embedded systems in these, a computer is used, more or less obviously, to control the operation and inter-working of a number of different devices used in a single process (or multiple related processes). The defining feature is that they use a hard disc to store information. Because these systems are more complex and involve many different components (any one of which may cause the entire system to fail), they are at greatest risk. Failures are most likely to be found in the application-specific software. About 20-30% of such systems have a fault: one in three of the faults would cause a significant failure (so that about one in ten of such systems poses a real risk to its owner).
Again, this shows that about 7%-10% of COMPUTER/SOFTWARE systems (we're getting away from "embedded chips" here, and getting back towards regular systems that can be fixed with a bit of programming) might be susceptible to crapping out. But not "burn down the house", you'll notice. Call in the programmers, and you could be up and running by 5.
Also note that this was written in *1998*, well before most major Y2K fix operations had gotten underway. This 1998 paper says that 7-10% of systems checked were found to be open to failure, BUT THIS IS BEFORE MOST Y2K FIXES WERE IMPLEMENTED. In case it has escaped your attention, there was a hell of a lot of effort spent in 1999, checking and fixing systems.
I think I see what may be a major reason for your mistake, though -- the paragraph you quote says, "poses a real risk to its owner". This does *not* mean "poses a significant threat to the survival of the company". In computer science terms, a "risk" is simply a door through which trouble *might* come, no matter how large or small. A "real risk" is one that has been identified as a practical, instead of theoretical risk. "Poses a real risk to its owner" simply means "is a recognized possibility of problems to the person who relies on the system". It does not mean the house is about to burn down. It means there's a possibility that your toaster might possibly toast only one side of the bread.
There is a ton of information here, which would seem to point toward major difficulties for business.
If you think so, you'll need to dig deeper, and try to present better examples than you've done so far.
This is where I based most of my conclusions about Y2K.
Ah, *there's* your mistake. And hey, didn't you just a few posts ago claim, "Suffice to say that I base my opinions on more than 1 source"? And yet, here you admit that "most" of your conclusions were based on this single, outdated, vague paper from the IEEE. Hmmm...
These people aren't alarmists, or yo-yos, or fearmongers, or hucksters. Something weird is going on. I really am curious how you can answer this.
Easily -- you've grossly misunderstood what the IEEE wrote.
Any further questions?
Actually, a VERY small number of DVD players MIGHT have a problem. One reason some of them MIGHT care what the date is, is because some movie rental places rented DVDs that were throwaway. They didn't have to be returned like a tape. They simply "expired" after a given date or so many plays. I believe the company that made such DVDs is now defunct (not sure though)… Another reason is some DVD discs use date codes to prevent it from being played prior to a given date. This allows the film distributors to make and distribute a movie before it is done playing at the theater. If the player's date system was screwed up it could well stop playing these types of discs.
Thought you might like to know…
I've been arguing here for some time about "embedded chips". Being an engineer that has used many controllers over the years in all kinds of things I had some idea of what I was talking about. That didn't matter to many here… I was just called names…
The bright side is, the proof is in the pudding… The pudding is pretty tasty these days… ;-)
Hey Dog you said Except I feel stupid. This is good, you're well on your way to recovery. ;-)
Take care.
You're getting more and more bizarre here.
You're trying your best to avoid rational discourse, aren't you?
You posted an article, I responded TO A CLAIM MADE IN THE ARTICLE.
Okay so far.
Rather than just let it lie, on the grounds that my complaint was with the article's author, instead of the person who posted it (you), you chose to get defensive, and tried to make it seem that I was being unfair for sticking to the very same point the article itself had made.
You responded to *me*. So of course I'm going to "defend" my position. Aren't you doing the same thing? And here's what happened: the author posted an example of his conclusions using company failure as an example. Yes, he did say Jan. 1. You're very right. But cripes, do you really think that he was basing it on just the .2% of businesses that happened to be operational on Jan. 1?
I came back and pointed out that if you're going to post an article, you should be prepared for criticisms of it, *and* should be familiar enough with your own article to recognize when I was addressing a point made IN the article.
Actually, you make one valid comment, usually followed by 4 or 5 disparaging remarks.
NOW, you're getting even more defensive, on even more tangential points, and whining about how you think I'm apparently saying that you shouldn't post articles you don't agree with.
Gee...imagine that. You respond, to my name, with your conclusions and insults, and then you are offended that I actually take the time to respond with my viewpoint?
For the record:
1. Post any damned thing you want.
Thanks for granting permission.
2. Learn how to accept criticisms of those articles without taking it personally.
If you don't want me to respond, then *you* quit taking it so personally and don't respond to me.
3. Understand that I never said that you should not have posted it.
I'm not the one that implied that I couldn't...but thanks anyways.
4. Understand that I never said that you had to agree with an article you post, *OR* that I even thought that you necessarily agreed with it (up until the point you got neurotically defensive about any criticisms of it, in which case it certainly *does* look as if you have a personal stake in the point being dismantled).
Wow. That was intelligent. Good one.
Suffice to say that I base my opinions on more than 1 source. So there. And suffice to say that I wasn't *talking* about your opinions, I was pointing out the errors in THE ARTICLE, and you need to stop being so strangely offended about that. "So there" yourself.
As I said, you post to me, I'll respond. I don't know why you're so "strangely offended" by that.
But hey, if you want to talk about your opinions, instead of the article itself (and your getting upset over criticisms of it), we can do that too. Earlier, you stated: "Here's what I expect: A 6 to 12 month of real economic hardship, precipitated by Y2K failures." Do you still stand by that prediction? Yes or no.
Yes I do. If I'm wrong, I'm in good company. But do I feel bad about it, or guilty? Hell no.
We could also discuss your wild overstating of the IEEE estimates, mischaracterizing them as "when a fireman tells you your house is going to burn down..." Shall we concentrate on that one? I challenge you to find anything in the IEEE Y2K pages that you yourself referenced, which concludes anything even remotely as drastic as "your house is going to burn down". I'll wait.
I'll grant you this to a point. I should have said, "when the fire department tell you that your house has the potential of burning down unless you do something different." They estimated that 1% to 7% of embedded chips will fail causing, I believe the phrase was "may fail in a way which would have serious business impact." You characterize that anyway you want, but it sounds pretty serious to me...and that's if just 1% fails.
Oh, and to answer the question that you've asked others twice in this thread, yes, I do know what the IEEE is, so I'll save you the trouble of asking for the third time.
Whew. Thanks.
Don't presume teach your granny to suck eggs, son. Yes, I'm quite familiar with the IEEE.
A simple yes or no would have been okay.
They weren't wrong. Your misunderstandings of what they wrote are what's wrong. Let's look them over:
So, to sum up, you're saying that my conclusions are wrong. Fine, I can live with that if it comes to pass. You have the benefit of drawing your conclusions on the IEEE data based on hindsight.
Any further questions?
Must...not...respond....way....too...easy...
Hello Douglas, normally there is a line for "Author" when an article is posted. When it is not specified it leaves it up to interpretation. When I initially read your post, I though you had written it. The way it is written it is an easy mistake to make (I think). I think this maybe why there was so much confusion on that point.
Regards, DB
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.
[
Top
|
Latest Posts
|
Latest Articles
|
Self Search
|
Add Bookmark
|
Post
|
Abuse
|
Help!
]
FreeRepublic , LLC, PO BOX 9771, FRESNO, CA 93794 Forum Version 2.0a Copyright © 1999 Free Republic, LLC |