In the Baker’s Dozen III thread, @Bicycle_Bill called Y2K a nothing burger.
But was it, really? I’m inclined to disagree; I think the reason it caused minimal disruption was because so many people and so much money was spent to combat it.
But what’s the Straight Dope?
The problem was overblown and there was also a lot of work to find and fix problems ahead of time. How it became overblown is a complicated story. The issue was known since the first time a year was recorded as only its last 2 digits but well into the 1990s almost nothing had been done about it. That fed the CTish fear of planes falling out of the sky and microwave ovens killing their owners, followed by the collapse of western civilization. It also fed the CEOish fear of lawsuits. Those things combined to provide the justification for widespread upgrade of hardware and software to identify and fix the Y2K issues, which had long been desired within the industry for many reasons. Some of those reasons were tangentially related to Y2K such as inconsistent storage of all time and date information, as well as incorrect leap year and DST changes. Other reasons were just the necessity of upgrading hardware and networks to meet user’s functional and performance demands.
I think you got it in a nutshell. If, hypothetically, nothing had been done to remediate it, Y2K would have caused a wide array of problems ranging from many minor ones to some really disruptive ones, but nothing on the dystopian scale that some were predicting.
I was doing consulting work for a large bank during this period, before and after we hit the magic 2000. It had nothing directly to do with Y2K, but we were hugely impacted by it because basically everything was. The paranoia was palpable, because so much of what banks do is based on calendar dates. In readiness for Y2K, all production software had been carefully scrutinized long before, the appropriate fixes applied and tested, and then everything frozen as solid as an Antarctic glacier. Any software changes or any new software during this Great Freeze had to be justified as a necessity and treated as a major emergency, and approved as such by a “Y2K Committee” comprised of pretty senior management well aware that a bad decision would put their jobs on the line.
The direct and indirect costs of Y2K planning, contingencies, and precautions for this institution must have been astronomical.
At the time, I was at a medical technology company; we made software for devices as well as higher-level enterprise type software.
We identified many, many potential problems and we busted our asses fixing them. Had we done nothing, the result would have been unpredictable, but it would not have been good, ranging from disruptive to catastrophic.
But we did do the work, and on the evening of Y2K, I was sitting in the office watching the support ticker and confirming we got through with only a handful of minor issues.
That hard work is why people with no direct experience or knowledge refer to it as a nothingburger, because we and many others made it so.
But perhaps we should have let things fail, because apparently people can’t recognize the seriousness of a crisis if it is averted.
This is a major sore point for me.
I believe this much was clear in the collective consciousness on the afternoon of January 1st, 2000.
When I was trained in programming at a major insurance company’s programming school back in the early/mid 90s, a large part of the curriculum was in COBOL (COmmon Business Oriented Language). At the time, COBOL was a bit of an archaic dinosaur of a language that had been around since the 1960s. It was also an extremely versatile and powerful language for writing programs that businesses needed, which is why it was still being used by companies 30 years later. When people were writing programs for banking, billing, the medical industry, payrolls, and you name it for businesses from the 60s through the 90s, nobody was ever expecting that the programs they were writing were going to be soldiering on when the year 2000 rolled around, or even that COBOL was going to still be a thing in the business and adjacent world. In the end, it still was because it was so good at its job. Which is why I was being trained in it to go through lines of existing code in a 30+ year old language to find and fix potential Y2K issues.
As it turns out, the job I found in programming had nothing to do with Y2K and I never used COBOL outside of school, but yes, a lot of time, money, and effort went into making sure that Y2K wasn’t a problem which is why it wasn’t. If such a huge effort wasn’t put into preemptively taking care of the problem in the 90s, Y2K would have been a disaster.
I was working in IT at the time, in NYC. We put in a lot of effort and money to patch, upgrade or replace systems/devices that were potentially vulnerable. In my opinion, the similar work done everywhere smoothed the way for a new generation of processes/policies around patching and upgrades that had a positive effect on businesses. So whether the sky would have fallen or not, it was a good wake up call.
Anecdotally, I was on call on 12/31/99, but attending a NYE party on the rooftop of a 27-story apartment building in NYC, near Times Square. Part of me was hoping I would see the city of New York go dark as the ball dropped a few blocks away. Alas, it was not to be.
Certainly the banking system would have been in deep trouble if the Y2K bug wasn’t fixed. Some testing showed things like 99 years worth of negative interest, which would have played havoc with personal and business balance sheets. I don’t think it was overstated at all – banking systems, government systems (probably including air traffic control systems?), accounting systems were all vulnerable.
As another engineer who was working on fixing a litany of issues related to Y2K I certainly believe that our work made the transition much easier. There would have been some significant issues, especially in finance, but I think the notion of planes dropping out of the sky was completely overblown.
I worked on a life insurance system.
The cost of doing almost nothing, and then working around the clock for a few days during which checks didn’t go out, would have been far less than the well-planned and tested effort over several years.
But maybe the ignore it and then do emergency fixes approach would have opened a door for a fraudster to exploit a period of chaos. Clean up from each fraud costs a lot when you have to make sure unlikely scenarios can’t happen again.
One overblown aspect is that there are other date cliffs specific to these systems. Many of our date calculations were based on the number of months since program inception, creating similar issues 1,000 months later.
Chicken Little = Henny Penny: The sky is falling! The sky is falling!
We avoided most potential catastrophes, even possible disasters, because of all the hard work and scrambling we did in software development in the late 1990s. I was a software test engineer back then, working hard with developers to flush out all the 2-digit years in our medical device code. And it wasn’t just in our code, but it was also any in the operating systems, communications protocols, and firmware that our systems ran on and with.
For someone to say Y2K was a nothingburger simply means that all of us that did the hard work did it all, or mostly, correctly.
Stanley Motss: It’s all, you know, thinking ahead thinking ahead.
Conrad ‘Connie’ Brean: It’s like being a plumber.
Stanley Motss: Yea, it’s like a plumber: do your job right and nobody should notice. But when you fuck it up, everything gets full of shit.
Stranger
Exactly.
My Y2K story: in 1999, I worked for a racial/economic justic nonprofit. It was a hot freakin’ mess, a small organization led by a charismatic control freak whose board president called her, at a fundraiser with hundreds of people (which lost thousands of dollars), a “Toxic personality…you think I’m joking! I’m not! She is!”
Anyway.
In July 1999 she hired a guy to work full-time on putting out a monthly newsletter. He dutifully wrote the first issue. She decided it wasn’t good enough and took it on herself, leaving him to come into work every day with literally nothing to do except lie on the couch and stress.
In like September, she told us of a meeting she’d been to where they’d warned that Y2K was disproportionately going to affect people of color and the working poor, and she instructed me to write an article for the monthly newsletter telling people how to survive the coming power outages and food shortages. I was the techiest person there, you see, so I should write the survivalist article.
I argued with her. All my tech friends were telling me the problem was well in hand, and I thought it’d be irresponsible to be alarmist about it. The working poor and people of color already had enough real stuff to be stressed about–why should we raise alarmist fantasies?
She suggested, not so subtly, that I was speaking from privilege and didn’t care about the folks we served.
I resigned soon thereafter.
But I was still on the mailing list.
When the newsletter came out, it did include an article on how to survive December 31, 1999, complete with tips on how to use a generator, how to obtain clean water, and how to grow food.
The first monthly newsletter was mailed in February 2000.
If humanity finds the will and the way to slow and reverse the course of anthropgenic climate change, will the concern have been overblown?
Same thing, although (imho) less dramatic impacts if corrective action failed.
Same with CFCs and the ozone layer, leaded gasoline, or acid rain.
They were problems that were recognized, effort was put in, and they were (mostly) resolved.
Problem is that then people point to the predictions of disaster that never came true as evidence of fearmongering, rather than the result of those efforts, and as an excuse to not tackle problems that have dire predictions for the future.
This level of fear mongering, I agree is overstated. But I would find it plausible (on a purely speculative basis, with no direct knowledge of the systems) to be concerned that parts of the traffic control network would misbehave and controllers would be faced with a lot of planes in the air and incomplete or misleading information about their positions, producing a crisis on how to get all of them safely to the ground.
For the record:
I was also involved in a low-level committee on this issue, and agree with all the comments above. There was no Y2K crisis, because of the long-term recognition of the problem, and the tremendous resources that were put into fixing it, so it would not be a crisis.
One of my friends is in software work, and he told me in early 2000 that he probably would never work again on such an important project, that had such an important result: nothing bad happened.
We fixed part of the year 2038 problem when we were doing Y2K work, but we recognized that we didn’t do a complete job. Some of those systems are still running and I expect them to be running for some years to come.