A question about embedded chips...

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

Please excuse my ignorance. I have heard and read much about the coming disruptions due to hardware, software and social failures. I am confused about something though as it relates to embedded chips.

Why haven't there been multiple high profile failures of these chips yet? I understand that the issue is around the date change so we should expect these problems to arise when their clocks rollover to '00. Why haven't these clocks begun to rollover?

I am assuming some type of distribution (normal or otherwise)of embedded chip rollover around the actual time/date of 12:00am on Jan 1, 2000. Certainly there must be some variation around what time these billions of chips think it is right now. Is this a bad assumption? If so, please explain. If this assumption is valid why then are we not hearing about the chips on the "tail" of the distribution already failing?

My pc at home cannot keep the time. I can change bateries and reset the clock only to have it be off significatly within a few weeks. I am assuming there is something wrong with whatever keeps the time (processor? bios chip?) Is this a reasonable anology to what I am asking above?

TIA, Brian

-- Brian Tanguay (btanguay@earthlink.net), December 06, 1999

Answers

It's not century date rollover yet!

---

Remember - Your 1999 Form 1040 must be postmarked

no later than April, 15, 1900.

-- snooze button (alarmclock_2000@yahoo.com), December 06, 1999.


Hi Brian,

Well, time's not up for embedded chips as yet.

However, at the end of 1996 embedded systems in two aluminum smelters didn't keep track of the date correctly (they "forgot" that 1996 was a leap year) and the day number in these systems didn't match the date in the controlling systems.

Result: the smelters performed an emergency shutdown. This caused the liquid aluminum in the smelters (and associated equipment) becoming solid. Both plants had to completely replace the equipment.

Since this happened in Australia, other smelters around the world using similar controls were warned in time to shut down their smelters before they became solid chunks of junk.

It's going to be interesting 4 weekends from now.

-- Dean -- from (almost) Duh Moines (dtmiller@midiowa.net), December 06, 1999.


I too feel that some problems should be rearing their ugly heads about now, in that Production lines are already dealing with "00" dates in their responsiblity for controlling stock rotations, etc. at most automated manufacturing plants. Oh yea...There was an incident where one such plant began manufacturing a product with a "00" out-date for recycling, and the automated processes caused the entire run to be put in the dumpster as 100 year old garbage. This was about 2 months ago. The problem has since been remediated.

I am unaware of the current language used in computer programing but it seems that the simple approach would be the best. Since computers only know truths as we tell them, how difficcult would it be to tell computers that "00" follows "99" with respect to dates?

-- Bryan (buddy2buddy@askagain.net), December 07, 1999.


Moderation questions? read the FAQ