They claim that they are ready. I'm hearing this from the electric companies, gas companies and the telecoms. But, let's focus on electric companies. I watched our local electric company last night on TV. They claim that the expect more problems with a winter storm than a "y2k glitch." Let's assume that they are telling us the truth.

Question: How did they do it when the Fortune 500 companies aren't done yet? How did they go through all those embedded systems? Do they simply have few date related systems? What do they have that the rest don't? Is there anybody on this forum that works for an electric company?

-- Larry (cobol.programmer@usa.net), July 29, 1999

Answers

Larry,

Did you ask your question at the euy2k.com forum? Maybe they have the answer?

-- y2k dave (xsdaa111@hotmail.com), July 29, 1999.

No, I'm not from there. Thanks for the info, I'll check it out.

-- Larry (cobol.programmer@usa.net), July 29, 1999.

It's simple really. The answer is in the term "Ready". Recognizing that it is impossible to fix it all, utilities have moved from becoming "Compliant" to "Ready", which means "Fix on Failure". "Ready" means they have canceled vacations, hired more staff and prepared contingency plans. They expect stuff to break, and they also expect to have enough people and resources available to fix it right away. I haven't heard anything said about the July 14 Executive Order forming the NIAC (National Infrastructure Assurance Council) Here is a link to an article no one seems very interested in: http://personal.riverusers.com/~aztc/August21.htm

-- Tim Castleman (aztc@earthlink.net), July 29, 1999.

I'm convinced that there is a lot of "type" testing and phone calls to vendors. That cuts down your testing time enormously.

-- James Chancellor, PE. (publicworks1@bluebonnet.net), July 29, 1999.

James:

I did see a reference to type testing by one of the "positive" electric utility engineers on Rick Cowles' forum a while back.

Although the IEEE came out *STRONGLY* against type testing as totally inadequate, this engineer's statement was, as I remember, "We were concerned enough about possible problems to go to the length of actually testing one of each type of device".

-- Jon Williamson (jwilliamson003@sprintmail.com), July 29, 1999.

The utilities are concerned that their vendors aren't prepared. In an insert in our electric bill, it stated that if things are to run smoothly, their vendors must be compliant. If they are not, they cannot guarantee that our power will remain on. They have hundreds of vendors--I hope the "mission critical," ones have their computers fixed.

-- not too sure (not too sure@nottoosure.com), July 29, 1999.

I think they can claim they are 'ready' because their CEO's have stocked enough wood for themselves to last 5 years. Of course, that's just one possible take on it...

Feeling a little cynical this a.m. ...

-- winter wondering (winterwondering@yahoo.com), July 29, 1999.

To Tim Castleman:

Thanks for the URL, I've followed the terms from "compliance" to "ready" along with the GPS, solar flares and so on. It's safer to assume that they're lying.

They ran a test on TV, for the public to see. All went well, which probably wasn't a real test. They showed the clock moving forward to 12/31/99 and nothing happened. But, tests can be fixed to look good.

If we do have problems and the grid winds up going down, they will have blood on their hands for calming the public so much when in fact they were lying.

I know what's involved in changing entire systems and I'm afraid of the many things that could go bonkers in 2000. It really irks me that the reporters and the electric company staff always have that y2k smirky smile on their faces when the subject is brought up. Is it really that funny? Millions of businesses at stake? Not to mention the lives that could be lost.

-- Larry (cobol.programmer@usa.net), July 29, 1999.

Winter, Cynical? Gee, I wonder why when the government hired spin doctors are supposed to keep you optimistic. You aren't one of those menacing hoarders, are you?

-- Mara Wayne (MaraWayne@aol.com), July 29, 1999.

Untl Rick locked things down, I followed the remediation of utilities as closely as I could. Several points became fairly clear:

1) Type testing is sufficiently valid to be acceptable, provided the "type" is narrowly defined -- same everything including current revision level of any software or firmware. Conversely, the argument that type testing is insufficient fails in general. Almost no product we buy has been item-tested, nobody is pre-sipping every beer or pre- lighting every oil lamp to validate it.

2) Vendor assurances of compliances are, as a general rule, verified by type testing rather than accepted at face value. Exceptions to this exist, both because some utilities are more trusting, and because if the vendor says no dates are used (and you can't see how one would be used anyway), there really isn't anything to test.

3) Information and devices are widely shared among utilities of a given type (nuclear, hydro, oil, natural gas). Accordingly, detailed lessons learned at one site are highly transferable to other sites. This has several ramifications -- you know (based on testing elsewhere) where the problems are, you know what doesn't use dates, you know what needs to be replaced and where to order which version. This speeds up the assessment/initial testing process tremendously, and that's what takes the real time.

4) Investigation (in great detail) determined that the number of *functionally significant* date errors was much smaller than originally feared. Once you combine a much smaller haystack with a predetermined location of each needle, your task becomes much more manageable.

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

Solar X-ray activity From maj.com

-- Test (Test@test.com), July 30, 1999.

Here's what NERC's going to talk about Aug 5-6 at their workshop in Chicago. Certainly somebody can arrange to be one of the waitstaff at the function?

It's a 2-page PDF and takes a while to load, for some reason.

-- lisa (panic@mongering.doomer_we're_all_gonna_die), July 30, 1999.