CompuServe Thread

I very much appreciate the thread… being in a bit of a rush at the moment, I'll download it and absorb it completely later. However, I did want to briefly acknowledge your reply about solids – Yes, I understand that boolean operations can be used quite effectively on many other geometries than solids – however, my point was that if 3DS is having such trouble handling faces properly, perhaps a solid modeling algotrythm would be a valuable option for geometires which simply won't work under the standard operation. I've spent the last hour trying different variations on a simple geometry that still won't work. I will send you the zipped file. Perhaps there's something very obvious I'm missing. however, it would be a shame to have to go through lot of work on a simple geometry because the boolean algorythm isn't doing it's thing. Again, I appreciate the thread which I will soon read, and whatever help you can offer. I will go for the asprin now… <g> Sincerely, Jerome Kalisz

Gary, Hate to be a nag, but please watch those quoted messages, and remove the # sign. Otherwise, those old message #s drive the readers and message sorters crazy, particularly if any replies get attached. Pretty please? Kevin Krell – Computer Support Associates

Yes, the hints G sent were valuable in understanding just how 3DS does its thing, and what the limitations are. Apparently, I assume, the problem with long, skinny faces is running into the mathematical tolerances… certainly, precision can never be infinite 😉 You hit the nail on the head when you said that what needs to be done is not a matter of reqwriting the manual and otherwise giving people more information to overcome 3DS' modeling shortcomings, but a re-tooling of the package and plug-ins to eliminate those problems for the users. Even if such limitations can be overcome by work and ingenuity, it creates inordinate amounts of work at the LEAST that needs not be. I would think the particular problem with booleans wouldn't require any major overhauls of 3DS' modeling engine at ALL. The basic engine is very powerful and quite apt on the machine level – as far as mathematical operations go. What is needed are some tools to manipulate shapes and modeling procedures. This could be part of the next release of 3DS, but could just as easily be done in an IPAS package (which would also save us the wait until 5 comes out). For example, IPAS routines that would, at the very LEAST, automatically adjust faces and edges to allow smoother booleans. How about an IPAS routine that forms a mathematical solid from a closed 3D set of faces for each object, performs the operation, and then outputs the result as faces, which is the language of 3DS? This would allow the user the option of performing a solid operation on an object that's just too difficult to calculate in faces, or would be to much work to adjust. You say you're not a programmer… neither am I much of one. nonetheless, I've seen the power of a simply written solid modeler that came from only 2 programmers. You say there's no better single alternative? You may be right, that no one has commerically wrapped decent features into a packag. However, if you use the power of the program I used in college for solids manipulation, and give it some easily created features, it would blow the doors off the modeling in 3DS without even trying. 3DS could *EASILY* integrate such a system with great improvements. I wonder, however, why it does not. I suppose 3DS has surrendered the modeling aspect to other packages, and thus doesn't even try to become a good modeler. However, even if you create your cad-accurate models in other packages, the abaility to perform booleans and freehand artistic enhancments on these models directly in 3DS would be a wonderful thing. As you mentioned, the limits in 3ds' modeling power has another element entirely aside from buggy booleans. And that's the inability to create shapes of any accuracy, even to the extent of a 100K shareware drafting program. You can't just key in a movement of a certain numeric amount for an object, you must rely on snaps and the display. You can't pick a point and draw a line at a certain angle for a certain distance, etc. Also, once objects and/or 2D shapes are created, you must manipulate the verticies themselves, instead of having geometries like spheres, cylinders, circles and lines existing as objects. Even if thiss ability were only present in the 2D shaper, it would mean a lot. As it stands, you *MUST* use a CAD package to generate shapes of any complex accuracy, and import them in… Ahh, well, enough ragging on 3ds' modeler. I just hope that the powers that be decide to be competetive with 3ds as a modeling package with a few simple enhancements, or a good IPAS routine or two. Heck, even a sharp programmer with some time on his/her hands could create an IPAS which would revolutionize 3ds' modeling. I speak from experience… 😉 Now, on to trying to find out how the things I learned about boolean manipulations in 3DS can be used to overcome the problems I've encountered…. That's the one missing part. Dividing edges coudl take days in itself for a simple rounded object…

Hi again Jerry, Oops. I should be more accurate when I exagerate <g>. The PRJ file was 550K (not 70K). A big difference from 8MB in any case… Mark

Hey Mark! How goes it? Some very tantilizing thoughts, actually. I'm certain that I would no longer feel frustrated by 3ds' inherent limitations with modeling if Acad could work well in conjunction with 3DS, and also repair some of the old Acad 3D modeling limitations (which I've heard it's done to some degree in 13). Perhaps the presence of Acad in the arsenal of Autodesk is much the reason 3DS never aspired to be much of a CAD interface. So… Acad is finally outdoing Jim Turner's little Gedit, eh? Hmm… it's about time 🙂 I look forward to getting a chance to work with it sometime. I know GEdit isn't exactly a powerful modeler as modelers have the potential of achieving, but that itself was part of the reason I was so amused… that a couple college professors had a program that elminated most of the problems that 3DS was notrious for. You mention booleans…. for most objects I like to create, there are few other ways to create them. BTW – a quick aside: 3DS & Acad as a combination would see its full potential if Acad could also import 3DS LOFTS, ass I haven't seen anything in Acad that compares to 3ds' lofting abilities. You're right about the overhead of solids. Which would suggest it not be the only, or even primary, method of modeling. However, knowing the power would be there if you need it would be a Godsend. You've probably read my other messages ssuggesting a solids IPAS routine that would construct a solid out of a surface mesh for the purpose of a modeling operation, perhaps more. You're good with IPAS – you might think of it, it would be an answer to many problems. You're also right about surface modeling being a mathematical miracle. Until I read the 3DS manual, I never thought it actually attempted it. I figured it did what I suggest in that IPAS routine…. generate a solid for purposes of modeling, then reduce it to faces afterwords. I look forward to seeing Acad & 3ds working together in NT, if not OS/2… perhaps when this particular combination hits, I will be in a possition that warrents getting Acad to go along with 3ds. That's one prohibitive fact; that a truly powerful package consists of the expense of BOTH 3DS and Acad. Hmmm… who knows, with the two heading closer together like this, perhaps someday they'll be 2 sides of the same package…. a package which you can get animation or drafting plug-ins as you needs them, but the graphic interface is actually the same thing. Many duplicated procedures could then be eliminated…. Anyways, FL…. things are getting along here. Working at an architectural firm as I try to find contacts to expand into architectural renderings/animations. It takes contacts, though…. Hope things are still going well for you in MI, as well… Take care and thanx! -JK