What exactly are you trying to achieve? Why does it matter whether the first process is blocked by input, stopped, waiting for a network resource, or simply taking time to calculate the output?

I'm thinking of making a "commander"-style program, and one of the things I'd like to do is display output inline in a terminal window if it's all generated up front, but automatically background the process and handle it in a separate window if it requires input (so that the user can go back to writing terminal commands immediately, but later interact with the interactive program).

I didn't really consider other cases for non-immediate output. I think I want to try showing the text inline first, wait patiently, but then move it if it reaches an input block. The primary goal is to handle interactive programs differently, automatically.

Of course, it might be different for cases like this, where my own bar program can spawn foo and connect to its IO streams. Maybe I asked the wrong question.