Also a sailboat (even a big one) is not stable and from experience satellite receivers can lose connection on slightly strong winds. So I guess this wouldn't work even if Starlink allowed it.
The starlink antenna may look like an ordinarry satelite dish -- but it is a super-advanced phased array antenna which uses a lot of signal processing to electronically steer the beam. This electronic beam steering should have no problems keeping up with boat motion.
From what I gather beamforming operates only in limited range, i.e. tracking satellites as they move small fraction of a degree per second. For bigger adjustments the dishes turn mechanically. It's not clear if beamforming could handle relatively small waves, but for big ones mechanical adjustment would be needed. I'm speaking out of my ass, but if I had to guess moving the antenna mechanically would not keep the link up in wavy waters.
The beamforming is fast enough. It can probably steer at a few kHz of steering-bandwidth, if not more.
Yes, mechanical adjustment for big waves, but that's a simple passive 2-axis gimbal and having the center of mass slightly below the gimbal axis crossing.
Where did they demonstrate it? Latest news was that they filed for a license [0] to install 5 user terminals on up to 5 Gulfstream planes - application that is still pending.
Speed of beamforming tracking is not the problem - what is a problem is the lack of compensation for movement of the base station (and necessary hardware for it) which is quite evident when people test it in motion [0] - which is quite understandable, SpaceX had to think about the features it actually needs to support to get current price point and use in motion is not the intended target for the current rollout. Tracking is much easier if you can assume that the ground station is immobile.
And no, putting it in a passive gimbal as some of the other commenters are suggesting is not enough to fix this as the direction of the antenna beam is not even close to being parallel to ground (like it is when a gimbal is used with a marine radar).
Plus there is the fact that it does not appear to be sufficiently waterproofed to begin with and the motor mount is rather flimsy (both a problem if you want to use it at sea instead of lakes...).
Marine antennas for Starlink will come, no doubt - but the current home user terminal one is not it.
This is absolutely true as long as they built the IMUs and control loops into the software to do that. My guess is by the time the service is available for moving users that will be the case.
i'm pretty sure boats (especially big ones) are one of the potential markets for starlink, so even if current receiver will lose a connection, there will be special receiver for boats eventually. Stabilizing phased array shouldn't be that hard, even without adjusting motors constantly.