"Didn't get good landing/impact video. Pitch dark and foggy. Will piece it together from telemetry and ... actual pieces."

https://twitter.com/elonmusk/status/553857574005915648

Don't military pilots say a carrier landing is much more difficult than landing on solid ground? Because the area is small, and not stable.

It seems like SpaceX have had to start out by attempting the most difficult possible landing scenario (for good safety reasons).

So they did OK, considering the difficulty.

Well, carriers have significantly shorter landing strips, require the pilot to either hit a cable or immediately take off again and usually move themselves to provide more favorable wind conditions to pilots.

SpaceX drone platform is easier in that respect, since the rocket is landing vertically and alot of those factors are mitigated. It's still a extremely hard and complex problem to solve - it IS rocket science after all ;)

It's an interesting comparison. At least, with a carrier you can go around, but I suppose that, for an aircraft landing on a carrier, the moment of irreversible commitment to landing maybe comes a second or two earlier than it does for SpaceX.

I don't know if the SpaceX vehicle is programmed to back off and wait if a big wave rocks the platform, but it could do that, wind permitting (edit: no, it couldn't, apparently, see reply by Denvercoder9). Fuel is limited, but it seems like it should be able to hover for a few seconds and wait for the platform to stabilize if necessary.

Anyway, they're not developing a sea-landing vehicle, so it's understandable if they don't spend so much time thinking about that. It's designed to land on solid ground, I think and the ground doesn't usually move, after all.

The thrust-to-weight ratio of an empty Falcon 9 is far too high to hover. Burn the engines too long and it'll just ascent again.

Their F9R test vehicle could hover. Isn't that essentially a mockup Falcon 9 first stage, but with only the centre engine?

https://www.youtube.com/watch?v=ZwwS4YOTbbw

I wonder how feasible / useful it would be to have realtime best-guess altitude from accelerometers / bouys / laser line-of-site relays etc to give the rocket up to date info to adjust thrust / vector to the drone-ship.

Or do they already do this?

Almost certainly they have really good GPS and inertial sensors on both. That should be enough. Any line of sight like laser or visual can't be trusted due to weather.

I doubt that heavy sea conditions are factored into design - I think currently the way to avoid that is simply not to launch when the weather is bad around landing and launch area. That's probably the cheaper and simpler solution.

I'd suspect the structure on a Falcon 9 isn't as strong as a F/A-18, relatively speaking. Also the F/A-18 engines don't provide downward thrust.

Wonder if an actuating deck would be worthwhile