Some thoughts....
PJ,
As we are aware, there is obviously something in the code that scasm will not compile. This could be due to the type or version of the program originally used to create/compile the model that is not being converted properly to scasm, or can be compiled with scasm. There maybe a switch in Ivan's MDLC or other MDLC method to get it complied, but I currently know of none. I have also run into some models that would not re-compile correctly after converting to scasm (even with no changes that I made to the code), but I do not recall having a line buffer overflow error when compiling to bgl, as you have encountered. Also, even if a model does compile there is a chance of some distortion sometimes that occurs. I have occaisionaly noticed with some models an overall size change and sometimes VC canopy distortions if the canopy is automated to open and close. MDLC is a very nice program, but I think there are limits to its ability to work with some models.
O