help! with Ardent Titan C subscript

[RAMontante]

[I apologize if this is the wrong newsgroup, but I can't find a Titan group]

I need some help with subscript-checking on an Ardent Titan C compiler;
it seems buggy to me (:-).  I enabled subscript-checking with some code
that I need to port (not mine), and it produces implausible errors.  The
executable generates lots (>10400) of subscript-out-of-range messages,
all for a handful of lines on which the referenced array has a constant
legal value, e.g.

	float cubeD[18];
	...
	cubeD[6] = other_array[i][j][k];
	...

(note:  the error message specifies cubeD, not the other array.)
Naturally, the output seems plausible, although I have no intuition about
its likely values, so I don't know whether there's a real error or not.

Am I missing something obvious?  I'll summarize any mail I get, if
there's interest.
Thanks,
    Bob Montante