Screwed the rest of the post supports in before it got too warm this morning:
then we ran around with the laser and measured the actual height of the surface the post rests on for each of them. I can then feed those measurements back into the theoretical post lengths from my model and adjust for the variation – the slab isn’t perfectly flat, and the upstand decidedly less so, so there’s a bit of adjustment to do for each post. This seemed to work pretty well on the shed, so I’m doing it again here.
Retreated to the aircon to play with my numbers.