Link

Good News, Bad News For Y2K Efforts - Gartner

Quote from Gartner's chief Y2K analyst, Lou Marcoccio, during a news teleconference this morning.

People are still trying to calculate the effects in terms of what will happen at midnight, Dec. 31, 1999 -- but that is not when most of the failures will occur, he said. Rather, people should be on the lookout for problems much earlier. The largest single "hit" will occur in July, 1999, he said, when many companies change to a new fiscal year 2000 and the Y2K bug arises.

Then in September, other companies switch to fiscal year 2000 with similar Y2K bug bite results, Marcoccio forecasted. In addition, problems will crop up in thousands of other companies that forecast their businesses only one quarter in advance, he said.

When all the counting is done, said Marcoccio, a full 25 percent of all Y2K problems will crop up before Jan. 1, 2000. Another 55 percent will occur during 2000 at some point, and another 15 percent will happen sometime in 2001.

"A lot of companies are ramping up their inventories this year to be able to deal with problems starting Jan. 1, and there are tens of thousands of contingency plans and active steps for that one day alone, but that's totally incorrect," added Marcoccio.

-- flb (fben4077@yahoo.com), March 22, 1999

Answers

Agreed. We go to Yellow Alert July 1. Shields up.

-- RD. ->H (drherr@erols.com), March 22, 1999.

flb,

That's true for accounting software. It's only partially true for other software. And most embedded systems that will cause problems will do so near January 1, 2000.

-- Linkmeister (link@librarian.edu), March 22, 1999.

I almost get the feeling that some of these pre-2000 dates are being set up as strawmen. While companies may experience problems before Jan/Feb/March, I have not seen a convincing argument that society will experience disruption before then.

Does anyone else expect societal disruption before the first quarter of 2000?

-- Puddintame (dit@dot.com), March 22, 1999.

Straw Men Indeed!! Accounting software does not run the world. Accounting software just looks ahead. And look aheads are easily fixed: just make sure that they don't look any further than 12-31-99. The real problems will crop up after 2000, and especially with the embedded chips. Of course nobody mentions that... GK

-- Glenna (kamoroff@hotmail.com), March 22, 1999.

Puddintame,

There are only a few reasons why there would be societal disruption before 2000. One would be if the stock market crashed.

Another would be if lookaheads in non-accounting software started failing regularly in December 1999. Another wild card is the GPS rollover in August.

Will there be noticeable disruptions due to fiscal-year rollovers in accounting software? I doubt it.

-- Linkmeister (link@librarian.edu), March 22, 1999.

If this is true, the fear will hit Wall Street. They will know before anyone and the accounting is everything to them. Don't think a market crash is impossible. And don't think it won't change our lives immediately.

-- Mike Lang (webflier@erols.com), March 22, 1999.

Puddintame, I know what you mean.

After reading some of Dr. Yourodon's articles I had decided in my mind those dates I considered most problematic. *None* of them have occured yet.

Then I would read an article (from mainstreem media) that would point out another date (such as Jan 1,1999) for potential disruptions. Normally this occured a few days before the date. I'd rethink my previous assessment of the situation, factoring in the new date. Inevitably nothing of particular significance would occur on the date and I'd be forced to rethink the severity of the other dates I had chosen as significant.

Then I realized what was happening. Yes, these dates *are* straw men. The media makes a hoopla about them, nothing happens, and then they say, well that means Y2K is going to be a "bump in the road".

That having been said, I'll come forth as someone who believes;

1. Nothing much will happen April 1 or 9. Those things that do happen will be so buried within their respective corperations/governments that their citizens/customers will not notice.

2. Nothing at all will happen Sept. 9. I have yet to see a convincing argument as to *why* this date is significant at all. Using a viable date as a null case seems absurd to me. 999999, yes. 090999, no. I predict that the media and the corperations using this date as a 'test' date for their systems will consider getting through the date without error to be a spectacular sucess.

3. Interesting things will happen August 22, Dec.25-Jan. 5.

There you go. I also believed nothing would happen when the Dow reached 10,000. Now if any DGI points to one of these dates and says, "but such and such didn't happen, so it won't happen for Y2K." Then I can just point to this thread. :)

-- Alison Tieman (fearzone@home.com), March 22, 1999.

One thing many of us are watching with interest is the Wall Street testing. From what I understand, the 12-31-99 to 1-1-2000 rollover simulation will not happen til early April. I don't know if it makes sense to think they may already have encountered significant problems in the early stages of testing. Whatever has happened, there seems to be no info available. Maybe they're just taking a low key approach.

If they do encounter big problems, we will not hear official reports about it anytime soon, but such crucial information would leak out quickly to friends and family of those in the know, to help them protect their assets. I pray it doesn't happen. There are all kinds of indirect problems the US financial sector could have that are totally beyond their control. But if they have major problems with their own software systems, it could cause a market crash and would bode seriously for all large software systems everywhere. So don't let up in your preparation efforts. We don't know when or if it will begin.

-- Bill Byars (billbyars@softwaresmith.com), March 22, 1999.

Participants in this forum reflect an incredible amount of ingenuity.

Although we may be subjected to very serious life threatening disruptions, let us not forget that the ingenuity and creative human energy here and elsewhere will not be dormant in the face of severe adversity.

Prognostications of an awful mess need not conclude in hopelessnes.

-- Watchful (seethesea@msn.com), March 22, 1999.

One reason Gartner considers July to be significant is their fixation on the financial / stock market community. Most large funds and institutions use forcasting algorithms for sophisticated financial instruments that look ahead 6 and/or 3 months. The U.S. stock market collectively reacts less than 6 months of forecasts provided by these algorithms.

One fear is that institutional programs that are driven without human intervention may make mistakes that translate into mistaken buy or sell positions. The trading systems will still be working perfectly this year. So the trading "judgement" mistakes will go through.

Other programs using the mistakes from one program trade could set the market on a spiral path in an hour before trading could be stopped. Can you see the effect?

-- PNG (png@gol.com), March 22, 1999.