18 December, 2009

Why Y2K Mattered ....and still does

I was flicking through a book called History’s Worst Decisions and the People Who Made Them and there’s a chapter in the back of it called The Y2K That Never Came. Ten years later seems like a good time to talk about a problem that is even more misunderstood now that it was then.

To recap, when computers were in their infancy, any amount of data processing that could be simplified was of great benefit, so when it came to dates, the year was abbreviated to the last two digits, as has always been common practice when writing dates. They knew even in the 50s, that this would cause problems come the turn of the century but assumed that by then computers would be so advanced that it wouldn’t matter. The problem with leaving things up to the future is that eventually, the future arrives. Computers were of course advanced enough by then but the software they were running was often built on top of decades of older versions that still expressed the year as two digits. In the mid 90s, programmers realised that they had better start taking the problem seriously.

And it was a very serious problem – not just one of having to recalculate the interest on millions of bank accounts. When a computer is programmed to assume the year is 19xx, that would mean that when the clock ticked over at the end of 1999, computers would think it was 1900. If a computer finds it is finishing an operation 100 years before it started, that’s an illogical instruction - like dividing by zero - and illogical instructions can cause crashes. That is where the danger lay. It wasn’t that people would get parking tickets for -100 years, it was that computers that controlled vital services may suddenly lock up. Anyone capable of reading this knows that when your computer crashes, it’s usually caused by something fairly trivial but you still lose everything else you were doing. So while your last crash may have been caused by that Cool Web Search toolbar that you didn’t realise you’d installed, that also caused you to lose the essay you were writing at the time, the email you were reading and the card game you were playing. Likewise, an illogical instruction caused by the absence of two little digits had the potential for knock-on effects that could have been disastrous in some cases.

As we all know, none of the predicted doom came to pass. That’s not because the threat was overstated. It’s because we dealt with it. Yes, pat yourself on the back humanity! For once, we actually saw a crisis coming and managed to avert it without mass devastation.

And because we managed to avoid disaster, the conventional wisdom is that there was never a threat to begin with and we were all worried about nothing. That’s the premise of the chapter in History’s Worst Decisions, because the decision it’s talking about is not the decision to leave the two digits out but the decision to do something about it. It lists the motivation as ‘greed,’ the culprit as ‘most of the western world,’ and the damage done as ‘hundreds of billions of pounds in wasted effort and stress.’ I don’t know where the authors get that absurdly round figure from but you can’t say the effort was wasted when it clearly worked. It’s one case where the experts are victims of their own success. Perhaps if we’d had a bit more chaos, people would be satisfied that there was a serious problem to be dealt with.

Much of the rest of the article focuses on the alarmist book, Time Bomb 2000 by Edward and Jennifer Yourden – a pretty easy target. Of course it’s true that there was some ridiculous hype over Y2K, with people preparing to go back to the pre-industrial age, but that doesn’t mean the threat was non-existent. The reality is that there’s a small minority of the population that secretly relishes the idea of stockpiling ten years’ worth of tinned food and bullets and taking to the hills. For them, any excuse will do, whether it’s nuclear war, Y2K or a Democratic president. The only thing they haven’t used as a pretext for indulging their survivalist fetish is climate change, which I find curious because it’s the kind of thing that should be right up their alley.

Whether the Yourdens’ book was an honest warning or shameless profiteering from hype, it’s right to look upon it as a quaint and amusing relic these days. But it should also be recognised that the absence of any calamities due to Y2K shows what can be done when we listen to the science, free from political and business agendas. In the increasingly unlikely event that we manage to avoid serious upheaval from climate change, “sceptics” will point to that success as evidence that that threat never existed either. No good deed goes unpunished.


Update: I'm happy to learn that not everyone has adopted the assumption that there was nothing to Y2K and that those who warned of it are getting due recognition. I received an email this morning with the following link:
http://lifeboat.com/ex/guardian2009

7 comments:

  1. This comment has been removed by a blog administrator.

    ReplyDelete
  2. I didn't know anybody thought it was all hype. I suppose it's better to avoid a disaster and have people say it was all for nothing than to take the disaster.

    ReplyDelete
  3. That was my thinking. If you have to evacuate for a fire or flood or something and the disaster never comes, that's a good thing, isn't it? If experts manage to prevent the disaster, that's even better.

    ReplyDelete
  4. Time to start thinking about the Y2K38 problem (http://en.wikipedia.org/wiki/Year_2038_problem), the point when the timestamps on 32-bit Unix machines will reset back to zero. Assuming that not all machines will be "so advanced" that the problem will just go away - reasonable given a lot of embedded systems run Unix - we could be in for some for fun.

    It's the End of the Unix Epoch as we now it.. and I feel fine

    ReplyDelete
  5. Oh, you prophet of doom, you! We survived the twitpocalypse didn't we?

    Now a new generation has the opportunity to become overlooked visionaries.

    I guess one of the unfortunate things about programming must be that if you do your job well, nobody notices.

    ReplyDelete
  6. *Sigh* Yes..

    A friend of mine worked at a large Chinese bank, helping them roll out their first internet banking solution. There was much fanfare on the day, red carpet and everything.. with the chairman of the bank set to be the first person to officially conduct a business transaction on the system.

    Shortly before the presentation Davo (not his real name) found a major bug in the software - one that would have prevented anyone even logging on to the system, let alone using it. Frantic minutes of coding were followed by even more frantic attempts at explaining why a new version of the system had to be deployed to the servers immediately. Grabbing one of the system administrators, in his best broken Chinese he hoarsely exclaimed "this..most.. important.. thing.. you.. type..password.. now!"

    Suffice it to say, with three minutes to spare, a new version of the banking system was rolled into place and the grand unveiling went off without a hitch. And no-one but Davo and one bewildered systems admin knew that anything had gone wrong.

    We are the janitors of the world. We fix the plumbing, we build the rooms you live in. We do all the hard stuff so you don't have to.

    ReplyDelete
  7. And for that, we trivialise your work, diminish your achievements and portray you as social misfits.

    It's a cruel world.

    ReplyDelete