CH10 (Government) question from author : LUSENET : TimeBomb 2000 (Y2000) : One Thread


The notion that Y2K might bring the government to its knees -- at the local, state, or federal levels -- may be a secret dream of libertarians, but it's such a mind-boggling concept that it's almost impossible for most of us to accept at a visceral level. We can think about it, talk about, intellectualize the concept -- but anyone who has grown up in the 20th century (at least in North America, Europe, or any other "advanced" country) has lived his/her entire life in the era of Big Government. To seriously suggest that all of this might change -- in much the way it changed for the citizens of the former Soviet Union -- is so mind-boggling that we don't really cope with it.

So the question is: at what point should we assume that it really IS going to happen? The interesting thing is that the OMB is requiring all federal agencies to file progress reports every three months; thus, we'll see quarterly reports all through 1998 and 1999.

The other interesting thing is that the OMB has moved up the deadline for agencies to finish their Y2K remediation work; previously it was 12/31/98, and now it's 9/30/98. The deadline for having all remediated systems "in production" is now March 1999.

For me, the "do-or-die" date is 12/31/98: if a significant number of agencies are still futzing around, without having finished their work, I think we can write them off.

Of course, the political pressure will be so great that there's a good chance that most of the agencies will declare that they're done, even when they're not! How would anyone know? Ultimately, the only way we'll know is by seeing whether their computer systems actually do work ... and that may not become evident until sometime during the year of 1999.

So, to put it another way: at what point will you give up on Social Security? At what point should you assume that Health & Human Services can't handle Medicare payments? At what point, if ever, should you take the risk that IRS won't be able to collect taxes or keep track of whatever taxes you have paid. At what point should you say, "Sayonara, Washington"?

-- Ed Yourdon (, December 26, 1997



I think you make a good point ... except that it's going to be impossible for the government (or any company in the private sector, for that matter) to "prove" that their systems are working. On the other hand, politicians are VERY good at offering assurances; Bill Clinton personally assured all of us back in mid-August that all of the Federal systems would be fixed in time.

Another part of the decision-making process that we all need to include involves the question, "At what point is it likely, if ever, that the government might institute regulatory restrictions or constraints to prevent us from making appropriate fallback plans?" The most obvious example involves the possibility of banking/finance restrictions if there is serious concern about the possibility of a run on the banks. But the other, really fascinating example involves the IRS: it depends very heavily on voluntary compliance on the part of tax-payers. A taxpayer who seriously contemplated the possibility of an IRS collapse might, for example, choose not to make any witholding-tax payments for the second half of 1999 -- on the assumption that by the time the IRS figured it out and came after him, the IRS would no longer exist. Obviously, I'm not recommending such an action, and intend it merely as an example. But the point of the example was this: if the IRS was worried about such misbehavior on the part of its erstwhile loyal citizens, at what point would it act to close off the loopholes that the aforesaid citizens might be able to take advantage of?

-- Ed Yourdon (, December 27, 1997.

For each agency, ask yourself: if they are going to fail, at what point do I have to start preparing so I'll be ready? That's the point at which I should assume they will fail, if they haven't provided some kind of convincing proof (not assurances) that they will be ready.

So it's a different answer for everybody, depending on what kind of preparations you, in your unique circumstances, would have to make.

-- Dennis Peterson (, December 27, 1997.

I remember thinking in 9/97 "gee, it will be great when some real progress is reported in early '98". I thought that would give me a better "handle". Unfortunately, that progess has not been reported. I downloaded and read Capers Jones' entire report. For an enterprise with a mid to large function point portfolio (read: all govt agencies) mid '96 is the most realistic "drop dead date" for repairs to be underway. According to the most recent quarterly report, social security completed assessment in 5/96. How convenient--we have an actual existing model to test Capers' hypothesis. The report (assuming accuracy here) shows that SS is about 70% done. Model supports Mr. Jones' hypothesis (which shouldn't be surprising--considering his background and expertise). Doesn't it defy all logic to expect successful completion by the other agencies (especially since they refer to SS practices as "best") in light of this data and the fact by the final qtr. of '97 most of them are barely beginning the actual repair process. Compound this with the government's exceptionally long and consistent history of bringing in projects late, over budget, and of marginal quality. Therefore, I cannot find any other logical conclusion. Very significant percentages of govt. mission critical systems will fail. Post y2k repair will be even more difficult (if not nearly impossible) after the fact. Even under "ideal" pre y2k conditions, the repairs are beginning in earnest about 2-3 years late. Can we reasonable expect them to be done by 2005? Mr. Yourdan, please tell me I've missed something here. I've made some gross miscalculation. I've applied some completely erroneous assumptions. See, my pysche still screams at me. I don't want to believe, in spite of considerable credible evidence that appears to withstand rigorous analysis. Alas, how can I logically reach any other conclusion.

-- p. Larson (, January 06, 1998.

If the Federal governmnet isn't going to make it on time, New York State has harldy started; then the local police will have major responsibility. I met with Sgt. Ed Kolor, who is responsible for the Town of Poughkeepsie Police department computer systems. They are retiring old computers, and becoming PC based, and upgrading those as fast as possible. I gave them a copy of the sw I had to test the real time clocks and my print out of the Ed's prepublication book, which they were very interested in reading. I told them about my intention of "buying the biggest shotgun I could handle" if I wasn't soon convinced that we weren't headed for disaster. They didn't like that idea any more than I do and pointed out that they not only had the responsibility for solving the problem but also had the added responsibility for not unduly alarming the public. I submit that we, who think we understand the problem should also take added care that we are not unduly alarming the public. That's not going to be easy as we work so hard trying to ferret out what's real from what we're being told. I was less than successful in convincing them there were potentially sever problems with embedded chips in their cars and trucks. Their comment was to "disconnect the battery for a few minutes and see what electronics had to be reset." Probably a good idea; perhaps not sufficient.

-- Art Scott (, January 12, 1998.

IRS LIKELY TO FALL SHORT OF Y2K FIX The Internal Revenue Service says its mainframe hardware and software probably will be Year 2000-compliant, but it hasn't developed a fix for its desktop PCs yet. Of the agency's 88,000 computer programs, 13,000 have been retired, and 40,000 have been fixed, leaving 35,000 scheduled to be upgraded by January 1999. A large percentage of the agency's mainframes are being replaced as part of the effort, but that still leaves about 1,000 mid-size computers and 130,000 PCs to bring into compliance. An anonymous congressional source says the picture is still pretty bleak, noting that IRS refund checks come from the Treasury, where "none of the mission-critical systems have been fixed yet." (TechWeb 27 Feb 98)

-- Art Scott (, March 02, 1998.

My view?

They willnot get it all done. They will not get all of the "mission criticial" work done on time. Some extraordinary exercises of central power will be attempted. But its a long way from Washington to Boise or Wichita or Sioux Falls. Something Boris Yeltsin has already learned.

In the states of the former Soviet Union there are many fundamental problems which impact upon the public order and economin advance: poor work ethic, lack of public confidence in the national government, lack of education and experience, outmoded business models, archaic technology, poor communications, lack of commitment to change at various levels of the bureaucracy, corruption, organized crime, etc. We can expect similar fundamentals to surface to a greater or lesser degree here if the intermediate case or worst case Y2K scenarios come to pass. Perhaps more on that later.

If the "reasonable best case", i.e. the "mission critical" systems come back on line in the first week, comes to pass then the problems will be the ripple effects on the economy and public administration of the failure of the systems which were sacrificed in triage. Look for something like the Great Depression of the 1930s, but without the stable family and community units or ethics which got much of America through it. Not pretty, but not the near anarchy of Russia today. Effective governance will be largely local. No large tax revenue flows to Washington, returned with strings. Civic administration can survive at levels considered acceptable or even optimum in the first quarter of the century. Large national programs (e.g. OSHA, EPA, welfare, Medicare) will be extraordinarily difficult to fund or administer during a 1 to 5 year period during which the electorate may well learn to live without them.

Your view?

-- steve pettit (, March 24, 1998.

I'm with Ed: the cutoff date -- for any organization beyond the small business level -- is 12/31/98.

Any sane IS/IT manager knows that for a project of this scope, you need at bare minimum an entire business cycle for debugging and testing. If new systems aren't in a position to go online at 1/1/99, they face having to at least cope with bugs and services outages.

Let me explain with a small-world example why a year is needed:

Way back in the days of MS-DOS 3.2, I used a version of

PC Tools for my file management needs (this was before PC Tools turned into an application suite, then a desktop under Windows 3.X, and later disappeared entirely).

Now, there were some differences in the DOS file allocation table in versions up to 2.something. This was a standard Microsoft practice and continues today.

Now, if you tried to use PC Tools with a non-compatible FAT type, you risked seriously screwing up your files. In order to avoid this PC Tools programmers put in a system call that checked the operating system's version.

Those of us in the outside world assumed that this logic when something like this:

     if (OS_Version < 2.X) exit program

What it was in fact was:

     if (2.X < OS_Version < 3.12) exit program

See the difference? The first logic causes the program to exit if the OS version is below a certain range. The second causes the program to exit if the OS version is below a certain number and if it's above a certain number.

Fast-forward to the day I install MS-DOS 4.1 on my system. There were no changes between the version 3.12 FAT and the 4.1 FAT, so PC Tools should run, right?

Wrong. The program logic tells it to exit with DOS versions above 3.12 -- which I'm now running.

So PC Tools exits for no reason other than a hard-coded call to the operating system.

This led to the creation of a program called SETVER.EXE, which is still in use even in Windows 95. It's used to "fool" a program with such a version call into thinking it's running on an earlier version of DOS.

So why is it you need a complete business cycle to test a new system? Because who knows where some programmer put in a call to the date as an expediancy? You might concievably catch all the major program logic that deals with calculations, you might catch all your databases and change the dates to include a century. But suppose there's some little program you use every day (like PC Tools) that on the surface doesn't look like it might make a date call -- but does so anyway?

What if this tool is run every three months? You might not find the problem until three months into 2000.

What if you have fifty programs that behave this way?

My advice: if you currently hold money or investments in any organization -- not just government -- that isn't Y2K-compliant by 12/31/98, liquidate what you have with them. Take it out and buy precious metals.

As far as the Federal government goes, I've essentially written them off right now. Perhaps they'll surprise me, but it's hard to imagine. But I'll be very public about it on 1/1/99. After that, there's no chance at all that they'll make it.

"John Smith"

-- "John Smith" (, April 22, 1998.

It was announced today that the US Senate is forming a Y2K committee. The article on Yahoo (Reuters) comments that various agencies have fallen behind in their reform efforts.

The more government committees involved in fixing the government's Y2K problem means the less that will get done as far as I am concerned.

Any thoughts?

-- Rebecca Kutcher (, April 27, 1998.

Rebecca: Definition of a committee (compliments of Robert Heinlein): An organism with three or more legs and no brain.

-- J.D. Clark (, April 28, 1998.

Somewhere in my readings of y2k, it was suggested that one of the signs of the, "bottom falling out" was the resignations of CIOs and heads of depts. In the the last 2 weeks, I have read that, 5 CIOs frm the DOD have applied for early retirement and the head of the OBM has resigned to become the head of Fannie Mae (a private corp) Add to this, IRS"Mr. Rosetti"s statement to be found in this discussion under the question,"If the IRS is scared, shouldn't we be?" I think we will all be seeing things deteriorate more rapidly than predicted.

-- Bill Solorzano (, April 28, 1998.

I felt a twinge of guilt upon reading your statement "The notion that Y2K might bring the government to its knees -- at the local, state, or federal levels -- may be a secret dream of libertarians,...." True, I have probably accepted the probable consequences more readily than I would if I were someone who accepted the status quo in regards to the ( Federal in particular ) government as acceptable. In fact, I have examined this particular quirk for months and decided it is ephemeral- just a lucky worldview to have in face of the impending Y2K "opportunity" ( a problem is just opportunity in workboots and coveralls ). Although a majority of Americans are currently, to some degree, "on the public dole", the loss of the feds to provide these basic services for any length of time would result in the necessity of local provisions. Assuming it takes the feds and/or states 6 months to get in line, sufficient local resources will have been activated making many of the programs supported by gov officials redundant. The results of a paralyzed government, even very short term paralysis, cannot be underestimated. I am a firm believer in Liberty. Nothing is free- someone pays. In our current atmosphere, most people think someone owes them something, and those who disagree are chastised and ridiculed by the Victims-Rights press. We need something to make us realize that we all have the power to choose today where we will be tomorrow, and that those decisions are sacrosanct, and out of the governments jurisdiction, despite their seeming folly. FREEDOM! Dale

-- Richard Dale Fitzgerald 2 (, June 03, 1998.

A related (and admittedly somewhat partisan) question is, Did anyone really notice when the Republicans "shut down" the federal government during the budget negotiations? Speaking for myself, I didn't experience the slightest inconvenience, and was more than a little disappointed when the federal offices were re-opened.

Once upon a time, we were proud of the inefficiency spawned by our "50 little experiments in democracy." I like the fact that people who like lots of social services can move to a state which provides them, while people who just want to be left alone on their land can move to a state which permits that. Federal encroachment only means a graying-over, a washing-out of America. For a President who nominally supports variety and diversity, he certainly seems to want a lot of unified services run out of his office.

-- Mark Zieg (, June 03, 1998.

Moderation questions? read the FAQ