The following link to a paper recently published by Mark A. Frautschi is perhaps the best written and researched article I have seen thus far dealing with the entire embedded system problem as a whole:

http://www.tmn.com/~frautsch/y2k2.html

At a minimum, this document should not be ignored. It addresses several key issues and provides and excellent bibliography and list of additional resources for anyone interested in researching the embedded systems side of this problem (i.e. all of us).

My thanks go to "No Spam Please" who provided the link to the Jim Lord article (http://www.y2ktimebomb.com/Tip/Lord/lord9841.htm) which in turn pointed me to the Frautschi paper.

-Arnie

-- Arnie Rimmer (arnie_rimmer@usa.net), October 13, 1998

Answers

Excellent reference. An important point is that the failure point for embedded systems will not be a homogeneous 1/1/2000. hat will be the peak date, but failures will start before, build to a peak and then subside over some YEARS post Y2K.

-- R. D..Herring (drherr@erols.com), October 13, 1998.

Well at least he admitted something I have known for some time - if you completely power down some of these systems and wait for the clock to reset - the internal absolute date will be reset to the default and the darn thing will function correctly for another considerable period till it times out again. Have seen this kind of thing a couple of times, really most chips of this sort were (IMHO) designed to be replaced every 5 to 7 years. So the first thing the tech will try is pulling the battery and powering down the unit, waiting a while, and then replacing the battery (if the unit has a battery backup) and powering back up. This is a common remedy for a lot of low level troubles - so at least you can say some of the problems will respond to a fairly common fix.

Note: this won't fix every problem of this sort, and I did not say it would. But it will fix a certain number of them. How many? Hard to estimate.

-- Paul Davis (davisp1953@yahoo.com), October 14, 1998.

The Frautschi paper has been around since early to mid summer. It is in fact, on an older thread under "embedded systems" on this forum. Once again, if you are going to read the Frautschi paper, you should also read the rebuttal to it. The Fraustchi paper is "fraught" with mis-information.

http://www.greenspun.com/bboard/q-and-a-fetch-msg.tcl?msg_id=00074G

-- K Golden (kgolden@solar.stanford.edu), October 17, 1998.

KGolden:

Thanks for point me to the rebuttal. I'm not terribly impressed but I'm still grateful you pointed it out to me.

It seems to me that if we get lost in a sea of syntax and symantics and name calling, we will find neither the truth nor the measured response that is needed. (You say 'compliance' I say 'ready', you say 'chip', I say 'system' and so forth). This issue is complex in the extreme.

The TAVA white paper was also 'fraught' with such errors of syntax and symantics and yet I believe both should not be ignored. The rebuttal, if anything, gave me higher confidence in both the Frautschi paper and the TAVA white paper.

I'm not saying that either one of these are 'the Gospel', only that the problem is extremely large and complex and ignoring it or 'claiming victory and going home' are not useful. Due diligence is required and preparation remains prudent.

Thanks again for pointing me to this information. I would be interested in any additional information you are aware of in this arena.

Arnie

-- Arnie Rimmer (arnie_rimmer@usa.net), October 19, 1998.