> The following was supposedly scribed by > john kosty > on Sunday 07 September 2003 12:33 am: >ust identified the fatal flaw in wanting to adapt >VRML: files are monolithic. This goes against the >whole design concept. Disregard last... }-( There is even more against VRML than that. Even if you can get over the monolithic file, the data is so dirty and yet rather useless for anything besides representation of the geometry (just a "pretty picture" as far as I am concerned.) The VRML spec is so flexible, that you can't write a program which will read any compliant VRML file without allowing for every flexibility written into the spec. If you haven't done a lot of programming, this won't be so easy to appreciate. As a simplified explanation, consider writing if-statements within if-statements inside a loop which is only activated if an enclosing if-statement is satisfied (yes, that is the overly-simplified version:) I have even found a GPL vrml viewer which utilizes a Perl library to load the VRML into a data structure, but haven't yet been able to understand the data structure (at least I haven't found it to be reliably any more predictable than the VRML that it has had parsed into itself.) This is basically only the first drawback to using VRML as an input format for modeling or engineering data. Now tell me how "triangle 72" knows that it is the left-front corner of the sink and where we store the make and model (or even just density) of the sink and tell me in a way that I can write a 30-line Perl script which pulls the data for every sink installed in every project done in the last ten years so that we can quickly and easily inform our clients that their sinks have been found by the state of California to cause cancer in laboratory animals. This is the sort of thing that should be possible with all of the data stored in all of the dwg files and other cad-system-formats which are kicking around out there somewhere. The issue is just how to get to it. I could write the program to obtain this data about the sink from dwg files, but it wouldn't be a 30-line script and it probably wouldn't even be as short as 3000 lines and even then it would miss every project which wasn't done in AutoCAD or didn't have a dwg file available on the search path for the final project drawings. --Eric -- The opinions expressed in this e-mail were randomly generated by the computer and do not necessarily reflect the views of its owner. --Management