Dave Winer decries lock-in, and Nick Carr effectively rebuts his argument. Here's what Carr says:

It may not be honorable, but as far as "ways to win" go, lock-in is actually extraordinarily sustainable - much more so, in fact, than features, performance, and price, which all tend to get neutralized more quickly than lock-in does. Many of the greatest franchises in the history of the computer industry, from the IBM mainframe to Windows and Office to HP's ink cartridges to eBay to the iPod and iTunes, have been sustained by lock-in. And that's going to continue to be true.

Printers are certainly a great example of this - vendors make them cheap, and then hook you on their cartridges for the life of the printer. They certainly make many times the cost of the printer on cartridge sales, and your choices are pretty limited - there's no such thing as a universal cartridge - even though such a thing would be fairly easy to create.

There's lock-in that developers and their managers deal with all the time, too - their languages and development tools. Once a shop picks a language, it tends to happily stayed locked into it - even in the face of evidence that another language might work better for some particular problem. Various forms of inertia make the developers happy to stay locked in.