3 ms·

The are always plenty of companies that don't belive a disaster will happen even with ample evidence. The fact that none of their systems caused significant har

• dspillett 2y ago
  > The fact that none of their systems caused significant harm strongly implies others over invested. Or that their systems were not truly that significant… One of the things people see, in hindsight, as overinvestment is large scale investigations (codebase reviews, archeology of documentation for embedded hardware, etc.) that found little wrong. Once it is known that for [whatever systemn] there were only a couple of problems that were easy to fix it feels like that effort was wasted: people stop seeing the risk of the unknown once they know. Also, all that time/money investment wasn't just in the latter part of 1999, many systems were updated before then because they were tracking due dates months or years in advance. Furthermore there were a lot of things fixed (or replaced) that weren't fully Y2K issues but things that would crop up later because offsets had been used on two-digit years (49/50 was a common roll-over point instead of 99/00, but that was far from universal, with the matter being compounded by systems that talked to each other). Some money was wasted. But I suspect, at least from a risk management perspective of knowing that things were going to be fine or mitigations were in place, that most of it would be considered well spent. Hopefully we've learned the true lessons (no, I know we haven't!): don't take shortcuts for temporary gain in production, and don't assume systems you put in place now will be gone in a decade or more so you don't need to design for longer.