>I am not sure what you mean here. These two do show as equal (tho with differing data widths). Can you describe what you are seeing, and what the difference is?
>
>Thanks.


well,
BINTOC convert the internal binary to a char variable,

CLEAR
* put 0 with different literals
* because zero is zero,
* conversion to binary must return 
* the same value

X=-0E0
? X,0H+BINTOC(X,'F') AT 20,0H+BINTOC(X,'B') AT 40 && here the bug go away because VFP convert the number to a Integer
X=-0.0E0
? X,0H+BINTOC(X,'F') AT 20,0H+BINTOC(X,'B') AT 40 && BUG 80 is the sign bit
X=-0.0E1
? X,0H+BINTOC(X,'F') AT 20,0H+BINTOC(X,'B') AT 40 && here the bug go away because VFP convert the number to a Integer
X=-0.00E1
? X,0H+BINTOC(X,'F') AT 20,0H+BINTOC(X,'B') AT 40 && BUG 80 is the sign bit

* another way
* this must return .T. because the conversion of the same value must return the same value
? (-0.0E1 = -0.0E0) = (BINTOC(-0.0E1,"B")==BINTOC(-0.0E0,"B"))

* ANOTHER EXAMPLE, INCORRECT INDEX WITH REAL VALUES

CREATE CURSOR BBB (AA B,RN I)
APPEND BLANK
REPLACE AA WITH -0.0E0,RN WITH 1
APPEND BLANK
REPLACE AA WITH  0.0E0,RN WITH 2
INDEX ON BINTOC(AA,"F") TAG T1
* EXPECTED : 1,2 OBSERVED : 2,1
BROWSE NORMAL