The way this model should have been done was on paper first, to a highly refined degree, then and only then in 3D. Just about any modeler will tell you not to try to design something like this at the same time you are building it, except for relatively simple “study models” where you’re trying to work something out that can only be effectively visualized in 3D. Maybe you don’t actually use paper, maybe you use a digital tablet or a 2D CAD program, but the principle is the same.
That's pretty much it, yeah... though I'm using a 3D CAD program, not a 2D one. (Most CAD is done using 2D sketches as the basis of features... though you also have the ability to do purely 3D geometry in most packages, most work is still done using sections which are revolved, extruded, swept, or blended into other sections. The use of NURBS and related tools is a lot less common, even when doing stuff like auto bodies... though I'm convinced it's actually a better tool for that particular job!
So why didn’t I do it that way? Mainly because when I started this whole thing I didn’t really know any better, and because at the time I was more concerned with developing my modeling techniques than I was with the design itself. As I have stated elsewhere, I have learned probably three quarters of everything I know about 3D modeling from working on this one project, everything from spline cage modeling to box modeling to some of the finer points of polygon editing.
Well, that's the fun part, really, isn't it? Try something you've never done before... find out it doesn't work... and struggle over it 'til you figure out what went wrong. Gives a real sense of accomplishment!
Right now I'm stuck in a similar groove with some advanced electromagnetic analysis tools which just aren't working the way I want 'em to... I'm beating my head against the wall over it. But when I finally figure out how to make the software jump when I want it to... damn, I'm gonna be a happy guy (and a much more marketable guy to boot!)
Unfortunately, if you don’t pay close enough attention to the shape of the room, it’s easy to paint yourself into corners, so to speak.
Yeah... been there, done that.
Actually, I have a similar problem with the Vega. I used a "mirror" operation in my secondary hull at one point (at the very beginning of my creation of the "interior sets") and there's an issue with calculation accuracy with that feature. So I'm having trouble properly exporting it. That's part of why I haven't done any more postings on that recently - I need to figure out if I need to rebuild the secondary hull from scratch, or if I can recover by replacing the "mirror" feature without causing everything downstream from there from collapsing!) I can export as a facet-based model, but if I try to send out a NURBS-based one, it simply fails... the nurbs file is big, so there's a lot of data there... but importing it into anything fails.
Thing is, I've got SOOO much work done on that, that I'm not really inclined to trash anything. All the other components (the saucer, the saucer's retractable nacelles, the cutter, the main warp engines, the engine pylon assembly) are all complete. But the secondary hull itself simply can't be exported! AAAAGGGGHHHH...