"No longer useful" is not a very good definition. I declare that dictionary obsolete! I would tend to define obsolescence based on the task. (Arguably this is

"No longer useful" is not a very good definition. I declare that dictionary obsolete! I would tend to define obsolescence based on the task. (Arguably this is just the "micro" definition corresponding to your market-based "macro" version.) Obsolescence is when another alternative exists for doing the task, that improves on the previous alternative to a sufficient degree that you prefer it, and don't see yourself ever willingly going back to the previous alternative. Just to pick some random domain, from outside computers: Steelmaking/metallurgical technology at the turn of the 20th century tended to produce a relatively high carbon-content steel, which was more brittle, especially at low temperatures, and could cause a ship to crack, take on water, and sink. As steel tech improved, and new ships were built with the newer steels, those failures started happening less often. Steels that were less "hard" but more "tough" made navigation safer. On the other hand, and maybe to your point, most of those older ships of the higher-carbon steel, that didn't sink, were still able to serve out their service lives. Maybe if you found a way to put them on routes in warmer waters, you could've extended their lives even more. Which, translating to your conception, would be adapting to the new market or sub-market. So hey, at least for shipbuilding, your model seems to fit! The idea of obsolescence reflects an often unquestioned bias toward, or belief in, progress. If that which is newer or came later is always better, and if we're always (or mostly) moving forward in a more-or-less unbroken arc of steady unalloyed progress, then of course older things are obsolete. When in fact it's usually more complex than that. For example when moving to a newer technology, there's often something else desirable that has to be sacrificed, or that gets ruined, in order to get the improvement. So progress is rarely unalloyed (no pun intended given the steel talk above), and obsolescence is also, rarely unalloyed.