WHat was Y2K?

The biggest Y2K problem wasn’t Windows or any of the other current operating systems. There were patches to deal with the date rolling over for all the current operating systems. The problem was that a lot of infrastructure relied on old code that was obsolete, unsupported, and unsupportable since the programmers who wrote it and engineers who implemented it quit or retired long ago.

Remember early electronic computers used vacuum tubes and only had a few KB of memory. Saving 2 bytes for every date by omitting the “19” meant saving a lot of time and money since it required less storage and more records could fit in memory at a time for faster searches and sorts. In the middle of the 20th century it was just too appealing to omit the first two digits of years and nobody then thought their conventions and even programs would still be in use 50 years later.

Many believers thought the world was going to end in the year 1000. In fact this idea was so common it lead to the term millenialist came to mean people who thought the end of the world (or era) was near. The same mindset came into play for the year 2000. To many the idea of a new millennium just seemed too momentous for nothing to happen.

The people who stockpiled food, water, guns, ammo, seeds, etc. thought that our infrastructure could fail when the year rolled over. If the power grid in one area failed this could lead to blackouts in a large part of our nation. If the banking system failed this could lead to a huge financial crisis. A large problem in one area could cascade to problems elsewhere.

Personally I didn’t think anything big would happen but I thought that there would probably be some inconveniences. I got a few $thousand out of the bank in case the ATM network crashed and I got a few cases of bottled water just in case something went wrong at the purification plant. I meant to buy a generator but I never got around to it. Y2K turned out to be a great thing for my industry (IS/IT) because it gave us the budget and time to patch, replace, and upgrade obsolete systems and software. Our budgets were huge for this. This did lead to a very slow 2000 because after updating everything in 1999 there was little maintenance to do for a while.

We fixed a lot of the Y2K problems ahead of time and far fewer systems failed than we expected.

We have faced and will face similar problems in the future. For a while it looked like we would run out of Internet Protocol addresses. When the 4 byte format was created nobody envisioned a world were nearly every business had a network and nearly every computer was networked. Add to that networked appliances and the need became even greater. IPv6 is largely unimplemented because we have found other solutions to this problem. Our address space was originally broken up into A, B, and C class networks that could accommodate 16 million, 65 thousand, and 256 addresses respectively. This wasted a lot of space since an organization may have needed more addresses than one class supplied but not nearly as much as the larger class. There were organizations with class A networks that had millions of unused addresses and ones with class B networks that had thousands left over. Classless InterDomain Routing (CIDR) blocks allowed organizations to have networks much closer to the size they actually needed. Network Address Translation also allowed much more efficient use of addresses since many devices can share a single internet IP address. Name based virtual hosts even allows many web sites to share the same IP address on webservers. Now every cell phone, fax machine, printer, toaster, fridge, car, and wastebasket can be on the internet. IPv6 also has other features but since we have dealt with IPv4’s limitations in other ways the adoption of v6 has been slow.

There are other problems looming. Now that wired phones, cell phones, pagers and fax machines all need phone numbers we eventually won’t be able to fit them into the 10 digits currently allocated. Going to longer telephone numbers will require far more software updating than Y2K did. The nine digit Social Security Numbers also will also eventually need to be expanded which will cause more headaches.

Hopefully now we understand that the systems we build have to be designed to be able to grow in ways we can’t currently imagine.