Ok, that OLE one is interesting. I have already sent a note to Ruud about that one. In its simplest form, here it is:

break on

getobject("WinNT://@LDOMAIN/@USERID,User")

exit 1

basically, an unassigned getobject. If you run this code with 4.10 you get a script error. If you run it with 4.02, you get the object address piped to the screen (which is what one would expect). That's the point I argued with Ruud but he said the former is expected (script error). Anyways, he mentioned that he would try and "improve that a bit" ... still looking ...

-Shawn