Y2K, one year latergreenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread
I happened to notice the other day that the long Y2K article I wrote a year ago (http://www.best.com/~mgoodfel/y2k.htm) still gets a dozen hits a day, despite the lack of any recent links to it on the net.
Re-reading it, I see that I assumed (like most of us) that there would be increasing concern throughout 1999, instead of the complacency we see. I also assumed there would be increasing numbers of visible failures due to systems using dates in 2000. And I completely omitted any discussion of the oil industry from my list of utilities. I talk about electricity a bit, but I simply didn't think of oil. Doh!
The description of the Y2K bug itself still holds. I still think it's the best long piece I've read that goes into technicalities. It could be made clearer for non-technical readers, but in my experience, they aren't interested anyway. They just want to know "what will happen?" I do more or less point out that problems will occur both before and after rollover.
So overall, I still give the article a "B".
I still get an occasional email asking me what I think of the Y2K situation so far. Recently, I got one asking me who I blame for Y2K....
My first impulse is to say "Who cares? Assigning blame won't make it go away!" But there are two good reasons to talk about it. One is that it would be nice if one group (namely programmers like me!) did not get scapegoated unnecessarily. Another is that in general, you can't learn from your mistakes if you don't look at them afterwards. We do need to understand where we went wrong on all this. Subtract the emotion, and that's the same job as assigning blame.
So here's my big picture summary of "Y2K, What Went Wrong":
In the largest sense, something like this was inevitable. Computerization has spread like wildfire, covering the entire world of business and government in less than 40 years. Politicians and CEOs have not really internalized an understanding of its strengths and weaknesses. They have no background in technology (politicians are frequently lawyers, and CEOs are finance and sales people) and no hands on experience (there are still lots of CEOs and members of Congress who've never touched a computer!) If it hadn't been this date-related flaw, it would have been some other technology problem. We simply would have pushed the envelope a little too far someplace.
For Y2K specifically, I blame several factors:
- Programmers really did not expect that systems thrown together in the early 70's would last 30 years. The industry was changing so fast, with each few years bringing new "generations" of computers and software, that it seemed ridiculous to plan that far ahead. You would literally have been laughed at for suggesting that these programs needed to work in 2000.
This was compounded by the peculiar nature of software. It can be infinitely reused, ported to new machines, run via emulation of old machines, etc. Without changes, it will continue to work practically forever, even as the documentation is lost, the authors retire, and the manufactureres go out of business. This is unlike other manufactured products which have a more definite lifetime. This meant that software could be taken for granted by many organizations, and greatly outlive their ability to maintain it.
The need for new systems to share data with old systems meant that a lot of legacy date handling remains. However, as 2000 approached, programmers have still been writing completely new code with 2-digit date handling. Jocelyn Amon has spotted programs on Java tutorial sites that are wrong! This is really sloppy.
So I forgive programmers for our reckless enthusiasm and poor planning, but there's no disputing that we did create the bug -- and continued to propagate it long after we should have known better.
- When people began to point out the problem in the 1970's, the U.S. government missed an opportunity to change its standards, which would have dragged much of the industry along. The Deparment of Defense vetoed the proposed change based on cost. There was already too much legacy software in use!
- The professional societies like IEEE and ACM also missed the boat. They could have raised awareness of this. Even this year, they could have done more. Why was PBS relying on a columnist like Cringley to make a Y2K documentary? Why wasn't the ACM in there giving expert advice, or even producing their own show?
- When the situation became very clear in the late 1990's, widespread ignorance of technology among decision makers, including company CEO's, led people to minimize the problem. Almost all remediation efforts got started late, and with insufficient resources. Organizations refused to make this a top priority, and refused to give it the extra time and resources needed to make absolutely sure that the critical deadline was met. As a result, a huge percentage of companies are still not done, just weeks away from the rollover.
- The US government also did not agressively try to determine the extent of the problem. For example, the Department of Energy could have picked a utility at random and done their remediation for them. Being government, they could have published a list of every item that had to be fixed without risk of being sued by all the supplier companies. This would have been a checklist and a warning flag to the whole world about how serious the problem was. Or, if nothing serious had to be fixed, it would have laid the whole lights-out question to rest. As it is, they handed remediation off to NERC, a trade group with no enforcement power. The details of remediation have been kept secret.
- Our legal system is too expensive, time consuming and unpredictable. Even now, information is locked up inside various companies by their fears of legal action. If you can't stand up and tell the truth without worry about getting sued and losing and paying damages, then there's a problem here!
- Political leadership decided that the definite threat of bank runs and stockpiling outweighed the (to them) unknown and unlikely threat of real problems at Y2K. So they put out the word that the problem was minimal, to avoid panic. Personally, I feel this will be a disasterous strategy in the event that the economic consequences are severe. They will have thrown away all credibility in order to buy a little time.
- The prevailing wisdom became that Y2K was no big deal. This view was created by government and corporate spin, by the overdone predictions of the doomers (including predictions about key dates that turned out to be no big deal) and a general inability to believe that a simple date-handling problem in computers could bring an end to good times.
Unfortunately, once this view of Y2K settled in, companies were even more reluctant to raise their hands and say they had problems (why be the only one?) And many smaller companies assumed the whole issue was overblown (even fake) and so did nothing. A year ago, I wouldn't have thought that what the public thought would make much of a difference (short of panic), but I was wrong. Public complacency has made the situation much worse than it had to be.
Y2K is already the most expensive disaster in history, and it hasn't even happened yet! Over $100 billion have been spent in the U.S., compared to ten+ billion for disasters like Hurricane Andrew or the Northridge earthquake. $100 billion is up there with the Gulf War or the Savings and Loan mess. Total expenses during 2000 will be far higher, in my opinion.
If we're lucky, and Y2K is primarily about disrupted supply chains and bad bills, the sloppy way this was handled and the happy-face reassurances given out will still leave everyone just a bit more cynical than before.
If it's an economic disaster, I expect there will be a severe backlash against the computer industry and technology in general. Hopefully, no violence! But given all the hype about the Internet and the techies getting rich off of it, I think we are really being set up to take a fall.
And the next time the government, the computer industry or "leading scientists" try to tell the public "it's OK, we've checked it out", they will probably get even more scorn than they deserve.
-- Michael Goodfellow (email@example.com), November 28, 1999
y2k is not the problem. It was a challenge that went unanswered by gov't and companies alike. The problem still is not y2k...its the ignorance and the denial of the people. Its the lies of the gov't. Taz
-- Taz (Tassie123@aol.com), November 28, 1999.
Y2K is a symptom of what else does not work right in our present evolution of civilization. The engine that is driving this manaically robust economy in the U.S. and elsewhere is built on a foundation of shakey values--philosophical, environmental, corporate, governmental, societal and cultural.
Hierarchical society with its small concentration of haves at the top and everyone else below has helped to cause this mess. This is not working. The urge to produce enough to make it to that elite tier at the top has helped to create a mess similar to what plagued earlier civilizations that crumbled from within. Only now we have moved beyond a geographical society crisis such as the Aztecs or Mayans faced into a global one.
Coming from a high-tech networking background, I understand what happened (good review of the why it happened by the way), and I agree with you that there will be a backlash towards the symptoms-- technology, and the people in the computer industry. Our friends, the legal profession, will probably emerge unscathed somewhat like cockroaches after a meteor blast.
A couple of weeks ago I would have argued that what needs to happen is a full scale "Manhattan Project" level change in the way the world stumbles through life. After reading Daniel Quinn's new book, Beyond Civilization, he has an interesting idea that we need to change society and civilization one family, one worker at a time. His views are a little bit similar to Ayn Rands in her book, Atlas Shrugged, by suggesting that each individual has to change the way his life is lived and how he views the world of work and the role it plays. Interesting and enlightening stuff.
Maybe if we did not create the Y2K issue, it or something else would have been invented to get our attention, slap us up the side of the head and help us to look at what else is going on.
-- Nancy (firstname.lastname@example.org), November 28, 1999.
Michael Goodfellow, excellent update, thank you for posting it here.
-- Ashton & Leska in Cascadia (email@example.com), November 28, 1999.
As time passes, it's getting harder and harder to explain both the lack of concern and the lack of problems.
So Goodfellow says we've spent over $100 billion, while Taz claims businesses didn't rise to the challenge! Goodfellow claims this makes y2k *already* history's biggest disaster. Yep, sure enough. So big that hardly anyone cares or notices. Instead, we have prosperity. But we don't *deserve* it, see? NERC, despite multiple, voluminous and detailed reports, have been keeping the *real* status "secret"! Right! The problem was minimized (to the tune of $100+ billion!), we started late, we didn't allocate enough resources. To support these slogans, we must dismiss the credibility of ALL status and testing reports. And when we do so, we just *can't understand* why we're not seeing an avalanche of problems! How mystifying!
But don't worry. If we're lucky, things will just be very bad. Otherwise they'll be catastrophic. Trust me. It'll happen. Later.
-- Flint (firstname.lastname@example.org), November 28, 1999.
Flint, from my readings, that amount of money has bought only remediation of some, *maybe* most, "mission critical" systems.
As Lane Core's recent Westergaard article states, we will soon know whether our definition of "mission critical" was correct or not.
As it stands, in my humble opinion, you seem to start from the idea that most or all business and government agencies had a pretty good handle on what needed to be done, and the money went toward doing it.
I suspect that, given your experience with software and the world in general, you have a much better realization of just how much "OH, SHIT" actually went on.............
Don't care if it is software, running a pizza shop, or building a deck behind your house. NOTHING goes as quickly or easily as planned.
-- mushroom (email@example.com), November 28, 1999.