[icarus_drowning]

As I understand it, the main difference is that the Falcon rockets are (or are eventually meant to be) much less dependent on refurbishment between launches. The Space Shuttle required extensive and expensive work between launches to the extent that many critics[1] claimed that it wasn't truly "reusable". (IMHO, while the shuttle program definitely didn't achieve its goals, it seems like calling it "reusable" is fair).

As for gliding back: the Falcon booster does not actually achieve orbit-- when the main engine cuts off, it is on a ballistic trajectory back to the ocean. While it may be possible to design some sort of gliding apparatus to "save" a booster on a sub-orbital trajectory, it is (again, as I understand it), much simpler to simply adjust that trajectory via a boostback burn that reverses or slows that trajectory, and to then perform a suicide burn[2] to recover the stage, either on a barge or (on lower orbit missions) back on a land-based pad.

I am not a rocket scientist, nor do I have experience in the space industry-- these thoughts are just based on what I've read following the SpaceX reusability program as closely as I can for the past few years.

I do play a lot of Kerbal Space Program, though.

[1] https://en.wikipedia.org/wiki/Criticism_of_the_Space_Shuttle...

[2] http://space.stackexchange.com/questions/10307/what-is-a-sui...

[SEJeff]

They also like to call the landing a "hover slam" in that it burns at the very last second and slams down pretty hard.

[ygra]

Ideally they reach zero velocity the instant the legs touch the ground. There was only one hard landing (that didn't result in a fireball) so far, and the legs are designed to take the remaining speed (they are replaced anyway and not reused).

They also have no other option. Even one engine at lowest throttle (70 %) is powerful enough to lift the almost empty booster again, so they cannot hover (which would make things a bit easier, at the expense of needing more fuel).

[WalterBright]

Yes, I learned that maneuver playing "Lunar Lander" in college. (The original BASIC version.) It was the only way to land without running out of fuel.

[verytrivial]

I can't find the reference, but I remember reading that the landing is complicated by the fact that even at minimum thrust, with the tanks nearly empty, the ship is so light that even with a single engine, it would accelerate back up after hitting zero velocity. So if the engines are restarted too early etc. it could "miss" the ground!

(If someone can confirm/deny this? I'm seeing approx 28T dry mass, thrust per engine of 66T and min thrust of 70%, but that dated guess work from http://space.stackexchange.com/q/4466/ )