@eval[2**64-1]

#1
The second one below ... I don't understand it. Shouldn't it be 0xFFFFFFFFFFFFFFFF?
Code:
v:\> echo %@eval[2**64-1]
18446744073709551615

v:\> echo %@eval[2**64-1=x]
0x7FFFFFFFFFFFFFFF

v:\> echo %@eval[0x7FFFFFFFFFFFFFFF]
9223372036854775807
 
#6
Hex values (like integer operators) are int64 (signed 64-bit) max.
Yeah, but 18446744073709551615 is 0xFFFFFFFFFFFFFFFF and treated an signed, that's -1.

0x7FFFFFFFFFFFFFFF is the largest positive signed int64, 9223372036854775807. I still can't figure out where it's coming from.

Code:
v:\> echo %@eval[18446744073709551615=x]
0x7FFFFFFFFFFFFFFF
 
#13
Apparently not too occasionally, or you would have mentioned it a few years ago!

I'm afraid you're on your own here -- you can write your own function to do the > 2*64 hex conversions. (But it will not be easy.)
I already have @CONVERT, which gets it right.
Code:
v:\> echo %@eval[2**64-1]
18446744073709551615

v:\> echo %@convert[10,16,18446744073709551615]
FFFFFFFFFFFFFFFF
 
#15
@CONVERT won't work with anything > 2**64, as it also uses int64 variables.
Well I don't know how it does it but @CONVERT correctly turns 18446744073709551615 (which is less than 2**64) into FFFFFFFFFFFFFFFF. And it turns 18446744073709551614 into FFFFFFFFFFFFFFFE. And so on. Maybe @EVAL's "=x" should assume a positive number is desired and use @CONVERT.