Dear Thomas, I am becoming addicted to your project!
I would love to come and be your assistant but I would need too many layers of thermal clothing at this time of year!! > And I could improve the script to handle > non-flat (not plane) walls, if you give > me a (simple) mathematical description > of it. OK, here is a mathematical description that is quite simple: All wall ARE flat, but only locally! You can divide your wall up into squares. It depends how unflat it is but sometimes squares of 500mm x 500mm will be OK. If the wall is very irregular, you may need squares of 100mm x 100mm. You have to obtain a LOT of survey measurements but, for each square, you estimate the local nodus height. You then use that nodus height for the little bit of sundial that goes across that square. Some squares will be blank so no problem. > ... the scaffolding will be removed after > xmas, so I have to hurry... I understand the problem. Scaffolding is expensive. I have TWO questions: Have you fixed the nodus yet? Do you get any sunny days at the moment? If both answers are yes, then you can mark a set of points where the shadow of the nodus falls on the wall at various times of day. Do this tomorrow!! It is important to put a sequence number against each point and the exact time of the observation. You then note the positions of each point on the wall. So, for each point, you have a sequence number, a time and a position. Then, back in the warm by your computer, you compare the actual positions with the predicted positions. You will see some irregularities which are just observational errors but if you see that all the points are a bit too low that that means the wall is leaning towards you a little. If just a few points are too low that means a local dip in the wall. If they are too high, that means a bump and so on. You can then go out into the cold and do some more surveying!! I envy you the project; I don't envy you doing it at this time of year!! Best wishes Frank --------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
