I would'nt say its 100% foolproof because if memory serves, whenever I "savesas" to change the name (not the first time) and then I did a compile more often than not the problem surfaced - so I make sure if I do a "saveas" I exit LC and come back in - which is why I went for over a year without it happening.
Just decided to test my hypothyesis ... and it turns out to be a wild goose chase.
I took a single stack with maybe 25 lines of code , did a save as and tried to compile a standalone. It came back with that useless error of "there is a stack with the same name" Save, Cancel , Purge - which always seems to go into a loop whatever I press, but no cursor problem.
I then did the same for the main stack with thousands of lines and multiple stacks where this problem has surfaced before, and guess what? It compiled without a problem. I tried a second time (I don't trust computers
) still no problems. Then again I think I have only had the problem once in the 9X series - it happened multiple times in the 6x series.
Sorry for the false "recipe"