Y2K To Cause 10% Of All Computers To Fail

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

http://www.newsbytes.com/pubNews/132622.html

A report just out from International Monitoring, a high-tech consultancy firm in London, says that its research has shown that all US imports will be badly hit by the Y2K problem.

Nick Gogerty of the company told Newsbytes that the firm's research has concluded that, along with a 15-day delay in all types of product shipments as a direct result of the Y2K problem, the Y2K issue is also likely to cause multiple and major problems for IT systems.

"Our calculations, which are conservative, show that around 10 percent of IT systems will be hit by the Y2K issue, to a greater or lesser extent," he said, adding that this is on top of an estimated 15 days delay in shipments on almost all products worldwide - a problem that will also be caused by the Y2K IT problem.

Gogerty explained the report's calculations to Newsbytes. He said that basic IT programming principles suggest that, when Y2K problems are spotted by programmers and the software is reprogrammed, around 15 percent of the reprogramming will result in further bugs.

"It's kind of bugs creating bugs, if you will," he said, adding that, even assuming that all Y2K first generation bugs are solved, then there will be a second generation of bugs amounting to 15 percent of the original volume of software problems.

"Our research has reduced that figure to 10 percent, which we feel is a conservative figure," he said, adding that the firm's research took in data from 140 countries around the world.

"Using this data, we built an economic model that integrated data on those countries' technological infrastructures, multiple PC environments and the economic profiles of the countries concerned," he said.

"We assumed a best case scenario and drew up a smokestack model. The conclusion was that, even with massive Y2K remediation, there will still be major Y2K computer problems, allied with delays in transport of all commodities worldwide," he said.

Gogerty, who is originally from the US and a veteran of the London financial services industry, as are his colleagues within the consultancy, said that the general conclusion was that there will need to be a 15-day stockpile of all goods into the US to compensate for the transport delays caused by the Y2K problem.

International Monitoring's Web site is at http://www.intl-monitoring.com .

Reported by Newsbytes.com, http://www.newsbytes.com .

16:42 CST

(19990628/Press Contact: Nick Gogerty, International Monitoring +44- 171-373-2856/WIRES PC, BUSINESS, LEGAL/Y2KBUG/PHOTO)

Copyright (c) Post-Newsweek Business Information, Inc. All rights reserved.

-- regular (zzz@z.z), June 29, 1999

Answers

Good story and post.

But .... I bet people will key in on "...10% of all computers to fail" and laugh "Hey --- thats nothing, no problem." After all, theres a 99% chance of dying (have to leave room for the rapture or assimilation by the Borg) and most people ignore it 'till on their death bed --- so denial will continue.

-- Jon Johnson (narnia4@usa.net), June 29, 1999.


Another outfit with something to sell

http://www.intl-monitoring.com/services.htm

Moreover - they do NOT specify as to whether that 15% figure is before or after testing is completed. Test suites are kept by all IT shops that do any large amount of programming. If it passes all the tests (with dates rolled up) the program is good enough for all ordinary uses. The idea that somehow major errors are going to be passed over just does not wash - not to anyone who has been inside a major DP shop.

-- Paul Davis (davisp1953@yahoo.com), June 29, 1999.


One question for the board: WHY have I never HEARD of these people before a couple weeks ago?

"Resistance is futile..."

-- Dennis (djolson@pressenter.com), June 29, 1999.


From the quoted company's website . . .

"International Monitoring is a specialist consultancy based in London which gathers, analyses and disseminates information about potential Y2K situations world-wide. The firm provides information for assessing the scope, scale and location of potential Y2K issues."

(Note: "potential")

So, lets analyse the news article quoted at the top of the thread, just for fun. . . (with a simplified analogy to begin with).

If "Fizzy Cola Corp" were to release a "news" article saying . .

"Drinking Fizzy Cola is thought to make you 10% sexier and 15% more likely to succeed in business"

Would you consider the information totally sound ? Or just a marketing device designed to influence you towards buying more of their product ? Clue: Would you expect them to state that . .

"Drinking Fizzy Cola has no scientifically provable positive effect whatsoever and in fact can in some cases contribute to obesity and tooth decay".

No. Agreed ?

Then would you not exercise the same judgement when reading Y2K "news" information originating from ANY company who's business activity is based upon selling services related to Y2K ? Would you go to a petrochemical company for impartial environmental information ? Or to a pharmaceutical company for impartial advice about vivisection ? Or maybe you'd ask a politician for impartial information regarding corruption in Government ?

I mean no disrespect whatsoever to this particular company or their products, as I'm sure they are a fine, upstanding, ethical and well intentioned organisation.

My point is simply that a vendor in any industry cannot automatically be adjudged to be an "impartial" information source as regards their own business environment. That doesnt imply dishonesty, just simple practicality and business sense.

Oh, and another note of interest. . from their FAQ page . . . (my emphasis).

"Q: How can you be so sure about damage costs and estimates?

Our estimates are based on the number of bugs, infrastructure available, technology in place and the preparedness of the individual country. The degree to which infrastructure is interdependent means modelling the risks exactly is impossible. We put forth relative risks and scenarios associated with them. International Monitoring can in no way insure that the scenarios put forth will reflect the scope or magnitude of the Y2K situation."

Nuff said.

-- Just a note (In@passing.com), June 29, 1999.


Of course it is impossible to quantify Y2k. It will likely remain so even after the fact. IM is attempting to do so. Their "guess" is as good as any, and is based on the best available data. If anyone would like to offer up their "guess", by all means do so, but prepare to back it up with your own research. You will be criticized as well. The fact is: no one knows, nor can they, before it plays out.

-- regular (zzz@z.z), June 29, 1999.


Another recent article about International Monitoring is on this thread:

http://www.greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=0010Wk

"U.S. Technology Imports Susceptible to 15 Day Y2K Delay"

-- Linkmeister (link@librarian.edu), June 29, 1999.


The idea that somehow major errors are going to be passed over just does not wash - not to anyone who has been inside a major DP shop.

Thanks: I needed a good laugh.

-- Lane Core Jr. (elcore@sgi.net), June 29, 1999.


Well, I guess 'regular' believes computers will fail. The actual story talks about systems (like accounting systems) rather than computers, but 'regular' doesn't notice. And the story says "IT systems will be hit by the Y2K issue, to a greater or lesser extent." Nowhere does it say 'fail' but 'regular' doesn't notice this either. In reading comprehension,'regular' is zero for two. If you believe computers will fail, who cares what the story actually says?

The fact that International Monitoring is selling a y2k service, and has a vested interest in spinning conditions to their benefit, has already been covered. "Spin" around here means minimizing or denying problems. Maximizing or fabricating problems isn't "spin", it's just the facts, ma'am. IM isn't exactly lying, you understand. The half- truth is a time-honored spin technique. If they told you how they made their money, their report wouldn't look quite so 'objective', so they leave out this little tidbit. Omitting critical facts isn't 'lying', quite, since they didn't actually make a false statement.

Now, how did IM decide the problems would affect 10% of systems? They use metrics developed from new developments. Are these appropriate to maintenance? Who cares, if we like the results. IM doesn't bother to mention this, of course. Another little omission.

For discussion, let's assume the metrics are pretty good (why not?). What IM does *not* mention is that many remediated systems have been returned to production after testing. The metrics say that 15% of repairs introduce new bugs. Testing finds and corrects most of these bugs, but IM doesn't say that. Another half-truth.

Once returned to production, many if not most of the errors that escaped the testing process turn up. The real world can be a damn good test. And indeed, we have good reason to believe that quite a few introduced errors have showed up in production, and we're correcting them as we go along. IM somehow doesn't bother to mention this either. Yet another half-truth.

Summary: Here we have an organization with a vested interest in exaggerating the scope of y2k problems, and makes their case based on at least four half-truths, never once having said anything provably false. And 'regular' takes this spin and misrepresents every bit of it in his title.

And rather than admit that their prejudices prevented them from noticing any of this, we find the doomers attacking those who *do* notice.

The TB2000 forum ambience results from this approach. It reminds me of Las Vegas, where you're surrounded by big prizes you might win, and every win is an ostentatious event with lots of lights and noises. The casinos are silent about the odds, and never mention that most people lose. They never lie, but they're masters at creating the wrong impression.

-- Flint (flintc@mindspring.com), June 29, 1999.


You must spend lots of time in Vegas, Flint!

Articles like the one above are 100% poppycock. Why would anyone base any conclusions on information from 140 countries around the world and assume a "best case scenario" in the first place????????? What are these people smoking????????

Getting back to reality...we still have 40+ percent of SME's doing nothing and nearly half of the fortune 1000 behind schedule. We have a government that tells us they are 93% compliant - unless, of course, you count software programs. We don't need articles like the one above to know we're in trouble!!

-- Don (dwegner@cheyenneweb.com), June 29, 1999.


Don:

They don't say they make a best case. You said that.

And what "information" do they have from these 140 countries? They don't say. If they said it was positive information, then of course you'd find fault, I imagine. After all, we have reason to believe many countries aren't even collecting information.

And what does "behind schedule" mean to you and me? You don't know, but you know we're in trouble, so it must mean trouble.

You make a case just like Froese does in the "stupidest thing" thread. Can't you see this?

-- Flint (flintc@mindspring.com), June 29, 1999.



>>They don't say they make a best case. You said that.

No Flint, THEY said it...

{snip} We assumed a best case scenario and drew up a smokestack model.

>>And what "information" do they have from these 140 countries? They don't say. If they said it was positive information, then of course you'd find fault, I imagine. After all, we have reason to believe many countries aren't even collecting information.

There IS NO positive information!! {snip} we built an economic model that integrated data on those countries' technological infrastructures, multiple PC environments and the economic profiles of the countries concerned

>>And what does "behind schedule" mean to you and me? You don't know, but you know we're in trouble, so it must mean trouble.

Behind schedule means that you are not in the end-to-end testing phase right now. Should have been there 1-1-1999. What's so hard to understand about that? Who is there?

>>You make a case just like Froese does in the "stupidest thing" thread. Can't you see this?

I reckon not!

-- Don (dwegner@cheyenneweb.com), June 30, 1999.


Don:

Your response to what I wrote is fair and reasonable. I should have done a better job of writing what I meant.

Let's start with those 140 countries. We know that our data on so many countries is poor at best. There are probably some indirect methods (sales of computer products within them, for example) we can use to estimate (within a *very* wide range) the exposure these countries face. As for remediation in these countries, our knowledge is nearly nil. Of course this doesn't prevent us from cobbling up an estimate, so long as we recognize that this is no more than a wild- assed guess.

Now we're very good at data analysis, no question about it. But the quality of the output of such analysis depends on the quality of the input. And there is little or no chance that the input IM used is more than a collection of assumptions, based on an absence of hard, reliable data.

You're correct, at one point they described their analysis as "best case". But before that, they described their assumptions as being what they "feel" were conservative estimates. This "feeling" took a couple of paragraphs to morph into "best case". Funny how that works, isn't it?

When you start to examine the basis for this "feeling", you run into what I described in a prior post on this thread. IM makes money from y2k problems. This inclines them to see potential problems. Their "conservative" estimates were based on several assumptions that I already described -- that no testing had been done, nothing had been returned to production, no fixes had been made to any problems that surfaced in remediated code, etc. We already know that these assumptions are contrary to observation.

When you are motivated to see problems being worse than they are, and you build known false assumptions into your model which have the effect of exaggerating problems, this is applied bias. Calling this procedure "best case" is at the very least self-serving. What can we suspect *might* have been the assumptions behind the fabrication of data from 140 countries from which little or no such data exist? Can you see a pattern here?

I'm not saying IM's projections can't come true. I'm just pointing out that the methods they use have a much better record for drumming up business than they do for accuracy.

As for "behind schedule", this (at least to me) isn't as self- explanatory as you imply. Worst (and to me, most likely) case is that many (not all) organizations won't have remediated and tested as thoroughly as they'd have hoped. In turn, this most likely means they will experience a higher error rate than they'd have if their schedule had been followed (or even followable at all). I don't doubt that the earlier an organization started (relative to the size of the task), the lower their future error rate. But please bear in mind that we're talking relative rates here. I firmly believe that the rate of y2k bug bites could *not* have been reduced to zero, regardless of adherence to any schedule.

The effect of these error rates on our daily lives depends on contingencies. What's contingent is the number of errors, the severity of errors, the recoverability of errors, and your own personal proximity to these errors (i.e. your bank, your power company, your employer, your telephone exchange, etc.)

The probability of any number of a very wide variety of personal impacts is unquestionably high enough to justify preparing against them to the best of our ability. Our efforts to determine just what that probability might be, are not well served by organizations that use self-serving half-truths and hidden assumptions to make their predictions. And it's hard to apply the same standards of credibility to those whose conclusions match our own, as to those with whom we disagree. Yet if we don't make the effort, we may expend our preparation resources unwisely.

-- Flint (flintc@mindspring.com), June 30, 1999.


Moderation questions? read the FAQ