One of the enhancements suggested at the Ann Arbor LCD Workshop was a way for Fast MC to read an .ini file and incorporate the geometry information from that file in its simulation. To simply store dimensions is straightforward, but for particle propagation Fast MC also has to deduce what components (at least all calorimeter components and the coil) are next to each other and where it needs to generate air volumes. This is of course implicit in the component dimensions, but would take some work to extract. Would it be acceptable for the time being to require that all detectors be of *type* small or large where * small-type detectors have the coil between EM and had, have the EM endcap inside the EM barrel; the coil length = EM barrel length * large-type detectors have the coil between the had and the muon detector; the coil length = had barrel length * for both designs the EM endcap sits inside the EM barrel, the had endcap sits inside the had barrel, and the muon endcap is outside the muon barrel (i.e., inner z of endcap >= outer z of barrel) It would still be possible to modify the standard Large or Small dimensions in the .ini file as long as it was done coherently. If anyone is seriously interested in using Fast MC for a detector which is, roughly speaking, topologically different from Small or Large -- that is, different in which components are adjacent to each other -- I'd like to know. Thanks, Joanne