> This kind of incremental development has been criticized a lot, precisely because it discourages scientists from trying anything audacious. Developing a sligh

> This kind of incremental development has been criticized a lot, precisely because it discourages scientists from trying anything audacious. Developing a slightly more efficient telegraph line every year still does not beat inventing e-mail. I’d be surprised to learn that it would have been possible to run packet data on the original telephone system. I’d be less surprised to learn that decades of incremental improvements ultimately enabled packet data to run on top of telephone lines.