Hacker News Reader

Comment by gd1 | Hacker News Reader

No, Y2K really was horrendously overblown. Is it too much to ask to have the panic and the response be proportional to the problem? You seem to be arguing that an overblown panic is ok as long as it achieves something. The panic can cause more damage/costs than the underlying problem.

View on HN

It wasn't overblown. You should see some of the systems large businesses use. Marriott still uses green screen terminals, for example, others use mainframes. ATMs often run archaic versions of Windows CE. If all air carriers shutdown because their archaic systems couldn't have handled the change over, then it would have been a disaster. It wasn't because it made news and everyone knew about it and remedied the systems. Clueless people sitting in front of modern mac books running modern OSes have no clue.

Mate, I'm probably older than you. It was 90 % overblown nonsense. Countries like Russia spent pretty much zero 'fixing' it and had no issues. Meanwhile the US had 'Y2K consultants', shysters selling $20 'patches' to fix your windows install, breathless documentaries about how your toaster and VCR would explode, planes would drop from the sky, the financial system would collapse...

Yup. I personally patched an air traffic control system that would invariably crash because of Y2K - in other words without the patch, there was no ATC... Some people miss this - the only thing we were doing with that date was showing it onscreen. But someone had written code into a fixed length buffer, expecting the year to only be two digits long, and that year was calculated as (current-year - 1900), so when 2000 ticked around, boom! Buffer overflow, system crashes and you're screwed. It went from being a minor bug to being a system destroyer.

The problem was real, and even if the patch to fix it wasn't rocket science, it still took time to track down the problems. Much of the affected code was more than 20 years old, and no-one had looked at it or even compiled it in years. Bringing that code back to life to be able to insert a fix was no small issue that could be corrected in the middle of a crisis. Multiply that by all of the other potentially critical systems, such as power transmission, medical equipment, factory plant, and you're staring down the barrel of a potential disaster. I don't understand why some people today want to pretend that there wasn't ever a real problem for Y2K - I saw failing systems with my own eyes.

In many cases it was overblown. We were forced to upgrade to Y2K compatible versions of pretty much every OS and 3rd party software we used. We also had to show that we were taking it seriously which, effectively, meant showing we were spending a lot of money on it. We spent about a week proving we weren't affected then spent about a year testing and shipping all the random upgrades. Ironically, the effect of all the change was much more risky than the original issue would have been even if we had been susceptible. On the plus side, we ended up with the latest versions of tooling etc which we didn't usually get budget for but it had nothing directly to do with Y2K.

Spoken like somebody who wasn't pulling 16 hour days remediating crucial financial infrastructure in the months leading up to it.

The panic in that case was not overblown in the tech community (I tend to ignore the panic in the general media because that's always overblown). The non-event of y2k was an amazing accomplishment, and replies like yours demonstrate just how under-appreciated it is.

Reply on news.ycombinator.com