Thank you all for your replies. I was looking at the map ctf-conquest (I imagine you are all familiar with it and its extremely well executed canyons) which uses triangular cylinders in the canyon area in a similar way to what you saw in my original posting (actually, the author subtracted the canyon shape with a series of triangular cylindrical brushes with normal, flat tops and bottoms, and then added triangular cylindrical brushes of the same shape, but with flat bottoms and varied tops to give the variety of terrain, whereas I just subtracted triangular cylinders with varied tops and bottoms to give the varied floor and ceiling surfaces), and I don't think that he did those by using subtraction, as was suggested, although I could be mistaken. The only way I can imagine that being done is with extensive editing of every vertex, which is what I was doing. All vertices line up between brushes in both ctf-conquest and my map to give a smooth, unstepped surface. However, obviously my map fails catastrophically on the bsp level. (I actually find my results ridiculous to the point of comedy.) Does tesselation fix this? I don't even know what tesselation is. I used cylinders with three sides (as previously stated), but perhaps if I were to take a tesselated cube and subtract half of it, that would give me the same results after vertex editing as the triangular cylinder, but without the bsp errors. Is that what the group is trying to tell me?
Ian
