>>Are you sure that all files are sharing the same globals file(s) and all
your
GLOBALS statements declared in the same order, with no other globals
declared before them?<<

Yup. Besides, subsequent behavior of the program confirms that nothing has
gone wrong with wholeInput$.

>>Have you recompiled the whole project to make sure that all the globals are
being mapped correctly?<<

Well, it doesn't appear to be a problem with the global variable itself, but
in any case recompiling alone (read: quit FB, trash pi file, go back in, add
carriage-return somewhere in GLBL file to make it think there's been a change,
recompile the lot) doesn't necessarily fix things. But adding a nonsense line
immediately before the trouble spot _does_ (presumably because the wonky
address is now occupied by code that doesn't do anything). I've had problems
of this sort even when the project consisted solely of a MAIN and a GLBL.

<< Do you mean that you printed wholeInput$ from within the same function>>

Yes, exactly. Well, if you want to split hairs, I called the text-displaying
function (which lives in an INCL file and is given a string parameter on the
spot, no globals involved) from within the troublemaking function (which lives
in the MAIN file, as does everything that MacsBug has ever chosen to finger in
a crash).

One of my testers recently asked whether an application's version number
refers to the number of bugs fixed, in thousands.