A quote from BBC's live feed on today's Philae landing, a post by journalist Jonathan Amos:
Assuming Philae can avoid the cliffs, the boulders, the fissures and the steepest slopes, it has a good chance of getting down in a stable configuration. But how do we know it's down? The action of the feet and legs touching the surface is to move a central pole running up the middle of the robot's main housing.
This will generate a signal that activates the screws in the feet and the harpoons on Philae's underside. It should also have activated the small gas thruster on the roof of the housing, pushing the probe into the surface. But, as we heard earlier today, we're no long sure this will work. So for Philae to succeed at landing, a soft surface will be preferable - something like a "snowdrift".
Philae hadn't yet landed at the moment Jonathan Amos posted his report.
Is it okay to say should have activated about an event whose timeframe is in the future at the moment of speech?
Can I say:
I should have visited you tomorrow, but I'm no longer sure I will be able to.