Hello again, I've now posted a second sample on this webpage:
http://www.gnu.org/software/3dldf/sundials.html#Calculating_Dials The PostScript and PDF files contain plans for vertical and horizontal dials for the latitudes of Goettingen, Germany (where I live), Chicago, USA (where I was born) and London, UK (for no particular reason). They also contain drawings in perspective showing how the dials were generated for each latitude. I suppose most of the members of this list already have ways of calculating dials, but if this were of any use to anyone, I could easily add more cities, as long as they're in the northern hemisphere. For ones in the southern hemisphere I'll have to think about what changes I need to make. The previous example illustrated the concept of using intersecting planes, but I actually didn't find the points this way, but rather by finding the intersections of lines with the various rectangles in the drawing. In this example, I do use an operation that finds the intersection line of two planes. (Just in case anyone on this list has used GNU 3DLDF before, which isn't very likely, the corresponding function already existed in the C++ code, but I hadn't made it accessible to users by writing a parser rule. It would therefore not have been possible to do this before yesterday.) The GNU 3DLDF code (in `sundl_02.ldf') could be neatened up considerably. For this sample, I just copied the code for each latitude. It should be possible to put it into a macro which would take four parameters: The latitude, the radius of the circle used for the equatorial dial, and two parameters for rotating the plane onto which the dial should be projected. There are probably other parameters that could be useful. It should be possible to use this code for generating dial faces for "skew" planes, i.e., ones whose normals don't point in the direction of a positive or negative main axis. However, except for the planes perpendicular to the equatorial dial, where the time division lines are parallel, there don't seem to be any that are of such general usefulness that it seemed worthwhile including them in this sample. One of the next things I want to try to do is generate dial faces for a cardboard model of a dodecahedron, along with the net for the model. This will be a good test for a routine for generating "skew" dial faces. Another thing I want to work on is finding the points on the time division lines where the end point of the shadow of a given gnomon will fall on a given date. I think I know how to go about this, but I'm not yet completely sure how to find the circle representing the ecliptic for any given date. As before, any comments, questions, or corrections would be very welcome. Thanks, Laurence Finston --------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
