Press still doesn't get it: "Mission-critical" is meaningless

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

3/26/1999 - Press still doesn't get it: "Mission-critical" is meaningless

THE Y2KNEWSWIRE.COM NEWS CAPSULE
The GAO has released another report updating the high risk areas of the federal government. It indicates that while progress has been made, there is still much to be done. Some major agencies are so far behind that they likely will not be compliant by 2000. Additionally, most sectors have yet to perform essential testing or to draft contingency plans. The reliance and system interdependency at all levels of government makes end-to-end testing critical. While the report discusses many agencies, it targets the FAA, HCFA, IRS and DOD as those with the most problems yet to resolve.

The Interior Department has announced that all of its mission critical systems are compliant. This is one of 5 agencies that has reported they will be ready by the March 31st deadline imposed by Clinton. That leaves three agencies that will not make the deadline including the Transportation Department, Health and Human Services and USAID. In the most recent OMB progress report, 79% of mission critical systems were compliant and that number is expected to rise to 90% before the deadline. [See essay below for more detailed comments on this compliance claim...]

The NRC released a statement saying they will be reviewing the Y2K plans for all US nuclear power plants. The reviews, to begin April 1 and run through July, will allow the agency to evaluate progress at each of the 103 plants. Last fall, the NRC audited the contingency plans of 12 plants and reported that it found no problems. Six more plants are scheduled to be audited in the coming months.

The Senate approved one bill that will encourage alternative to lawsuits in the event of Y2K computer problems. The bill will allow a 90-day grace period in which problems can be fixed, and it will restrict damages. Most are in favor of the bill in order to keep lawsuits to a minimum, but Democrats worry that consumers will not be adequately protected.

The Fed has stated it is concerned about its ability to provide emergency loans to banks in the event of computer crashes. Currently banks are able to borrow funds through a discount window, which could increase dramatically should one or more bank computers fail. This would leave the Fed in a difficult, if not disastrous situation. Therefore, they have asked for legislation that would allow them to make the loans without the usual restrictions. Certain types of collateral could be counted that are not at the present time.



THE PRESS STILL DOESN'T GET IT: "MISSION-CRITICAL" IS MEANINGLESS
Essay by the Y2KNEWSWIRE.COM staff
Federal Computer Week, a publication that usually adheres to excellent reporting standards, slipped up in a big way in a March 24 story entitled, "Interior joins elite federal Y2K compliance club."

The first paragraph says, "The Interior Department announced today that all 90 of its mission-critical systems are Year 2000-compliant." But the second paragraph says, "Interior joins five other agencies that have reportedly fixed all computer systems weeks before the March 31 deadline imposed by the Clinton administration."

The subtle -- but all-important -- shift from "mission-critical systems fixed" to "all computer systems fixed" is a fatal mistake. Yet here's a well-respected magazine equating the two terms; basically implying that "some" equals "all" and completely missing the point that the Interior Department isn't even claiming full compliance of all systems.

Story at:

http://www.fcw.com/pubs/fcw/1999/0322/web-interior-3-24-99.html

WHAT DOES MISSION-CRITICAL MEAN?
It means anything the agencies want it to mean, of course. It could easily mean, "Any system that is compliant." Y2KNEWSWIRE.COM has maintained that moving some systems into an ambiguous "non-critical" category and then refusing to fix them is a recipe for disaster. Peter de Jager claimed this, too, years ago. And he was right: planning on fixing just some of the systems is a losing proposition. Why? Because companies will save the most difficult systems for last.

The federal government has not published any standards describing a rigorous definition of "mission-critical," by the way. And now, as the March 31 deadline approaches, federal agencies and members of the press are taking "mission-critical systems" to mean all computers.

It's yet another example of naive optimism: hoping that mission-critical really means "all the computers" because that's what they want it to mean.

THE SCIENTIFIC METHOD HAS BEEN SCRAPPED
Remember the Scientific Method? It requires a burden of proof before something is accepted as fact. And the stranger the apparent theory, the more proof is required. Yet for Y2K, the Scientific Method has been all but scrapped. Claims of total compliance are accepted wholesale with no proof whatsoever. In fact, claims of partial compliance are accepted as claims of total compliance. And that, of course, is exactly what we're seeing here in the Federal Computer Week story.

But remember, when the deadline was originally set, it didn't say, "Fix some of your computers by March 31." No, the deadline was for all systems to be remediated.

Yet now, we're not even getting promises that the "some" will be ready. The Clinton administration reportedly expects only 90% of the "some" to be ready by March 31, 1999. And strangely, this Federal Computer Week story adds, "The Environmental Protection Agency, the National Science Foundation, the Nuclear Regulatory Commission, the Small Business Administration and the Social Security Administration are the other agencies that have reported that all their systems are Year 2000-compliant."

Really? The NSF is believable, but the NRC and the SSA? Get real. The NRC is only talking about internal compliance, not the compliance of the nuclear plants they regulate. And Social Security has been touted by the Clinton administration as being compliant since late 1998 -- even as the agency itself was saying it wouldn't be completely ready until July of this year.

SO WHO DO YOU BELIEVE?
Until these agencies come out and define "mission-critical," claims of mission-critical compliance are meaningless. It's exactly like saying, "All category five systems are compliant." But when you ask them, "What is category five?" ...they don't know. It's whatever they want it to mean. They might as well say that all Floopsie XG systems will be compliant.

At what point did the requirement of compliance change from "all" to "some?" And at what point did the press lose its ability to distinguish between these two simple words? Here's the answer: at that same moment they accepted the mass delusion that Y2K would be a 72-hour bump in the road. One delusion reinforces the other, creating a self-reinforcing fantasy world where computers really don't have to be fixed to achieve compliance. It's a seductive fantasy world, but it's not the real world.

DEFINITIONS:

all 1. The whole quantity, extent, duration, amount, quality, or degree of; the whole; the whole number of; any whatever; every; as, all the wheat; all the land; all the year; all the strength; all happiness; all abundance; loss of all power; beyond all doubt; you will see us all (or all of us).

some 1. Consisting of a greater or less portion or sum; composed of a quantity or number which is not stated; -- used to express an indefinite quantity or number; as, some wine; some water; some persons. Used also pronominally; as, I have some.

Clearly, some does not mean all, but in exactly five days, you're going to be bombarded by positive-sounding statements from federal agencies that are begging you to believe "some" does mean "all." They are going to subtly use slip-of-the-tongue language, starting out by saying, "All our mission-critical systems are fixed," then shifting to, "All our systems are fixed." It's precisely the same slip of the tongue now used to make people believe the United States has no more national debt. Politicians use "national deficit" and "national debt" interchangeably, impervious to the fact that the two are radically different.

PROVE IT: FIRE THE PROGRAMMERS
You know how to tell when the remediation is complete? Simple: the day after they fire the programmers. Until the programmers are fired or moved to non-Y2K projects, the remediation effort is not complete, and claims of compliance are just hot air.

For every federal agency that's planning on claiming compliance, we have a challenge: fire your Y2K programmers or contract team. Then, we'll believe you think you're done. That's a step in the right direction, at least.

###

Link

-- Arachnid (no@way.com), March 26, 1999

Answers

Arachnid::

Nuce catch of someone's excellent analysis.

Yo, Norm?? How is this spin???

Chuck

-- Chuck, a night driver (reinzoo@en.com), March 26, 1999.


hey, chuck- drop me an e-mail.

-- Drew Parkhill/CBN News (y2k@cbn.org), March 26, 1999.

Funny thing, there are a lot of resumes floating around from people with Y2K credentials who are having to retrain to find new jobs. Sound like there might be some projects finished?

-- Paul Davis (davisp1953@yahoo.com), March 26, 1999.

How could that possibly be, Paul, given that under the best of circumstances those "finished" projects would be undergoing testing that would require those IT folks to still be around in the event that problems are uncovered. And then there are all those "non-mission critical systems" that still need remediation....

A more likely explanation is that the projects never got off the ground, and the Y2K Fix-It charade is now being replaced with the Y2K Contingency Plan-It charade.

Y2K CANNOT BE FIXED!

-- Jack (jsprat@eld.net), March 26, 1999.

Jack -

Not true. Many of the coders on our internal project have been freed up (after three+ years before the mast) and are now working on other projects. The PM just shipped out to the state capital (where things are in very poor shape indeed.) Our test teams use different personnel from the code teams (and some of the code work was even outsourced overseas). It's bad practice to have programmers test their own code, and besides, most of the coders would rather die than get "stuck in QA". It's OK - the QA folks think the codejockeys are all a bunch of primadonnas... 8-}]

There will be lots and lots of completed Y2K projects between now and year end, and lots of folks available from them.

"What?", I hear some Tonstant Weaders muttering. "Has Mac come over all pollyanna?"

Hardly. I just think that we'll end up with some form of a "Pareto result": 80% complete. Which means 20% not complete. Which means 1 in 5 organizations miss the deadline. As in 100 of the Fortune 500. As in some very important parts of state and county and city governments...

Which is why I'm preparing for some serious bad times...

-- Mac (sneak@lurk.com), March 26, 1999.



This isn't binary. Both Jack and Mac are correct. Many projects never got off the ground; many did and are being completed but it won't cover the needed ground to prevent a very bad result. As usual, Paul is covering up a shallow analysis with a pretense of expertise.

-- BigDog (BigDog@duffer.com), March 26, 1999.

Assuming 'mission critical' has some useful meaning (other than 'whatever we end up getting fixed'), the distinction could be important. I might have missed it here, but my interpretation was that failures in critical systems would have public impacts (and possibly cause domino effects), while failures in non-mission critical systems would not.

My take is that if ALL mission-critical systems are perfectly remediated, and NONE of the non-mission critical systems are even begun, we'd suffer a fairly short period of inefficiency and possibly some minor resulting economic slowdown. Some minor delays, some unimportant services not available for a while. Nothing to really worry about at all. Mission-critical failures can make life very hard for some innocent bystanders.

Is this definition way off base?

-- Flint (flintc@mindspring.com), March 26, 1999.


Oh, Good Lord:

Did everybody not notice this is from Y2Kloosewire? "Gosh, what a reputable source" LOL! how many of you have looked into what 'other' sites might be tied in with them? hmmm?

Mutha, signing off for the weekend, but looking forward to more entertainment from the "GI"'s.... A HAH HAH ha ha!

-- Mutha Nachu (---@awayfor the.weekend), March 26, 1999.


Moderation questions? read the FAQ