Often, a newer technology (or even a non-technological system) will be better than its predecessors, but the switching costs will be high enough that it isn't worth moving away from the older one. If you work at a large, old company, you may well work on a software system from the 1980s or before, and it probably frustrates you sometimes, but it may not be worth the company's while to get a new system up and running and tested, all of the existing data ported over, and you trained on it. This is even truer if there are network effects -- hydrogen-powered cars might be better than gasoline-powered cars, but until enough people adopt them that it's worth putting up hydrogen filling stations, it's not in any individual person's interest to switch over.
Monday, June 06, 2011 :::
All of that notwithstanding, if you had told me 15 years ago that we'd be testing IPv6 in 2011, I would have been surprised that the costs of using IPv4 hadn't surpassed the costs of switching at least ten years earlier.
::: posted by Steven at 9:46 PM