> As an > example: code file A RUNs code file B which has an error near the end such > that you may not notice that it didn't really complete its execution, but > control returns to code file A which picks up after the RUN. RBG7 is still > aware of an unresolved error and, upon the code that attempts to EXIT, the > RSHOW does not get executed.
My problem is not on exiting, it's on entering. Sometimes, the application never gets to the point where anything at all is displayed -- in this particular application the user sees a couple of DIALOGs (which are shown) to get name and password, and then nothing until the database is connected and the main form shown -- and we never get there. In the occurence I saw at my PC, it was literally a case of typing a CONNECT statement at the R> prompt and going into some kind of loop in which R:Base would wait for a time out, let me know it happened, and then wait again. This is obviously very installation/environmentally dependent because it only occurs on a few machines, and not that often. -- Larry
