The Millenium Bug

@Clockstops I was someone who fixed programs that needed fixing at the time. It wasn't some kind of scam to get paid work.

Small businesses would have just needed to apply any updates to software that they were using, and that would have resolved nearly all issues for them. Any business with custom software or specialised software would have needed to run a project to check and then fix that software. None of this is made up or a scam.

Computers weren’t changed from using YY to YYYY in 1999, the issue was recognised much earlier than that so most things had been designed to use YYYY well in advance. It was older technology and new technology which was designed with a link to older technology, which needed to be rewritten and tested. Most home electronics were designed with that in mind, and runs independently anyway.

I know we spent a long time telling customers what a disaster it would be if they didn't employ someone to fix everything. Most did, but those that didn't didn't seem to have any issues.

Former colleagues of mine all worked incredibly hard on this so that there were no issues, it was a massive undertaking and the fact that it is now being commented on as ‘something of nothing’ is insulting to the sheer amount of work they put in.

I worked in a bank at the time and we were on edge wondering whether cash machines would still work, would people be able to access their accounts and funds etc? I can only imagine what things were like across other industries. I remember us all breathing a massive sigh of relief that everything went smoothly (thanks to those hardworking software folk)

Why are you arguing with people who actually know what they are talking about 👀

Yes, but in those days most of the software was downloaded on the machines, so the engineers could only fix what they were asked to fix.

People using computers weren't the people who needed to fix things. It was the software they ran on. Developers would have been issuing upgrades and stressing the need to install them etc.

@Clockstops like a PP has said, there probably were some minor issues, but nearly all businesses ran projects to check their systems and fix them. So any issues were minimised, and small, not newsworthy. People were also on call to fix any more serious issues quickly.

Not everything was tested though. Only the things that were included in the project. Things like the small electronics used in homes, like PP GM's weren't tested and nothing failed.

@Clockstops If you google you will find that there were indeed a lot of things missed. Thankfully most were minor issues like bus tickets with the wrong dates but there were a few more serious ones like nuclear reactors having problems. This is why those of us working in the industry were on call 24/7 so we could fix things there and then.

It certainly wasn't created deliberately, the pioneers of computerised systems didn't all get together and decide to write programmes whose date fields couldn't cope with a new century with unknowable impacts on eveything

I was due to give birth around that time and I was told to bring blankets and a torch to hospital if I went in on that date. So that was quite scary. I held on till 9 January to be on the safe side 😬

Yes, but only if you tried, and surely there would have been countless people using computers in small businesses etc who did nothing and yet were still able to carry on as usual.

I know what the issue was, but I find it difficult to believe that everyone fixed everything, and yet there were no problems at midnight.

It wasn't overhyped, it was an issue which was identified and which people worked hard to solve

I think it is quite possible that nothing of world news worthiness was missed.

Software developers are pretty thorough people, with a policy of code checking each other's work, testing etc. The whole industry was very aware of the problem.

It was a quite 'straightforward' thing to find and check for.

Things were missed, absolutely they were, but we had people in call all night and issues were sorted quickly. These were small things though, luckily. We had whole offline systems running tests for months before the actual millennium happened, 99.99% of things were already picked up in those simulations. We didn’t just hope we’d noticed everything, things were tested.

@Clockstops it was a genuine issue. As PP have said it was caused by older programs storing dates with two digits for the year so YY instead of YYYY. This was a totally reasonable decision at the time, it was down to having very limited memory storage in computer systems at the time. It was a reasonable assumption to think that those programs and systems written in the 60s, 70s etc would be replaced/upgraded by 2000.

If no one had fixed the issue, then as soon as those older systems had to deal with dates in the format 01/01/00 meaning 01/01/2000 not 01/01/1900, they would have produced errors and either behaved unexpectedly or stopped working altogether. I can imagine your banker colleagues would have been unimpressed to have to deal with that happening in lots of systems all at the same time at the start of the new millennium! Much more sensible to plan in advance to check all systems, and fix any that had the potential to go wrong.

Yes, but absolutely nothing did happen and whilst an awful lot of work was done to remedy it, surely it's not possible that nothing at all was missed?

It was a massive industry. I was working for a bank at the time and the money that was spent, both by the bank and by business customers on the bank's advice/insistence.

I'd to know to what extent the problem was created deliberately and what exactly would have happened if we'd all done nothing.

In a nutshell, when computers were developed their memories were small in comparison to what they are now, so most dates were stored as 2 digits, dropping the 19, so 87, instead of 1987. So when midnight happened dates would’ve snapped back to 00, instead of 2000. This meant formula based on time would stop working, and since nearly every formula has something in it which uses a time calculation it meant it would stop working. Eg satellite technology would’ve been kaput, you can imagine how catastrophic that would’ve been.

As many PPs have said, working in IT at the time, we spent a long time preparing for it to make absolutely sure everything ran smoothly. Our company was a banking software provider and tbh the previous years introduction of the Euro was an even bigger problem for us, but no one seemed to realise how much of a potential issue that was. 🤷‍♀️

I think it was something to do with computers being unable to calculate the time that anything would take because they would think it was 1900 instead of 2000, so for instance, a plane itinerary which started on 1st January 2000 would confuse the computer into thinking it had been in the air for 100 years and completely be unable to cope with it! I may be wrong though!

If the computer system thinks its 1900 and not 2000 then it may potentially be looking at the wrong information when making other decisions. So possible that air traffic control might not be showing today's information, whatever that might need to be - maybe weather, flight schedules etc etc.

Software crashing when encountering a negative number but needed a positive one (eg this year minus last year) is an obvious one.
Software crashing (ie just giving up) in an uncontrolled way is never good.

Anything checking how old people are: this year 00 minus year of birth 55 means you are negative 55 years old. Crash!