I like taking the train (as opposed to flying) when I head to NYC. This might come as a surprise to people who know me, as they know that I'm fairly skeptical about any expansion of high speed rail in the US. The thing is, it works in the Washington - NY corridor, and in the NY - Boston corridor. There are other places in the US where it either works now, or could work - the region around the Great Lakes comes to mind, and also the coastal California corridor between LA and SFO. That's not to say there aren't problems though.

If the railbed isn't ready for high speed trains, the infrastructure cost for laying those down is enormous - especially compared to planes, that don't have to stick to a restricted space. On longer haul routes, rail makes even less sense. In the abstract, I'd love to take a comfortable rail trip across the US. In reality, there aren't enough people who want to do that on a regular basis to justify it - why take a multi-day train when you can fly in 5-9 hours (depending on whether it's non-stop or not)?

To relate this to software, I think it's another case of picking the best tool for the job. When I have to head to NYC, the train is a far, far simpler and more efficient option than air travel. When I have to head to Cincinnati? It's a different ballgame.