It is very much out of my head, the plan tends to change quite radically up until I actually start cutting metal - it is more fun this way. And as soon as I actually see a part I immediately think of a better way of doing the next step.

I am not entirely sure of the stiffness of the 50mm square Al beam with the supported rail on top, but it probably isn't bad. I think a bigger problem is fighting it turning into a parallelogram which is why I made the plate which is sitting on the top in the photo. If I started again I would have made one of the Square Al beams 5 inches longer so I could put a good triangular cross bracing plate overhanging the end.

Progress is going to be slow for a couple of weeks, as it is the Science Festival here in Cambridge and I have many many other things I have to build for it.