The intent is for this to be overridden via MDB, and I thought it'd be a bit more obvious that you should use /W and /X if I made it an int.

If the intent is that it should be overriden by mdb then it shouldn't be static. That gives the compiler leeway to inline it as it believes it shouldn't be modified by anything outside the source file.

I did check the program text and test the toggle, but I see your point.  Do you want me to remove the static qualifier from the other variables above this one, too?  They're called out as being for debugger "tweeking" also.

There is this issue that static or not static does not seem to affect "visibility" of variables for mdb. I think it should be considered bug:)

When most of this code was written compilers would never inline globals that were considered statics. Now, some compilers will. There have been many cases when transitioning from studio to gcc4 that the compiler would opt to inline the variable instead. I disagree about mdb -- if a variable is in a symbol table, mdb should be able to tell you that regardless of whether it's global or local scope. If the problem is disambiguation and knowing which one that refers to in a multi-module system, that's something we should improve. When you're debugging you still want to be able to see what a symbol's value is.

Yes, thats what I meant.