LUT5 probably BUG!

Hello...

Fidling with my gantry homing I use LUT5 function. If I understand all of that correctly then I can declare it's buggy!

lut5.1.in-0 = TRUE
lut5.1.in-1 = FALSE
lut5.1.in-2 = TRUE
lut5.1.in-3 = FALSE
lut5.1.in-4 = FALSE

with function set to 0x330 they return TRUE (correct)
with function set to 0xe10d they return FALSE (should be true! as 0x330+0xdddd=0xe10d)

I was checked that with halmeter and function number and inputs are ok, just result not!

sadly found that even calculator at wiki.linuxcnc.org/cgi-bin/wiki.pl?Lut5 is bugy.
See atached schematic for lut. It's divided to 4 sub parts all joinded with two or gates at the right.
I used only 4 input pins and pin 5 (in-4) is tied to false.

So As I calculated LUT by hand (few minutes ago I find calculator ) for the 1'st term is only inverter. And as input0 should be 0 and other's are dont care the result should be sum for:

3 2 1 0 weights

0 0 0 0 0x1
0 0 1 0 0x4
0 1 0 0 0x10
0 1 1 0 0x40
1 0 0 0 0x100
1 0 1 0 0x400
1 1 0 0 0x1000
1 1 1 0 0x4000

sum of 0x5555

but python lut5.py -n4 '(~i0)'
give me sum as 0xffff wich is totaly wrong.

all other parts seems to be ok.

and I tested individula parts and they worked ok, but when it's joined the result is wrong!
and here is calculator formulas:
python lut5.py -n4 '(~i0)' #Wrong result - should be 0x5555
python lut5.py -n4 '(i0 & i1)' #0x8888 is correct
python lut5.py -n4 '((~i1) & i2) & (~i3)' #0x30 is correct
python lut5.py -n4 '((~(i1 | i2)) & i3)' #0x300 is correct

and all of them with final OR
python lut5.py -n4 '(~i0) | (i0 & i1) | ((~i1) & i2) & (~i3) | ((~(i1 | i2)) & i3)' #0xffff is wrong result

It's not so bad....

I find that there are nothing wrong with LUT, but manual is not correct in some terms.
As I had lut divided to 4 subparts and every subparts works then I added function to join them together.
And it's works in most cases, but not all! The weight's should be ORed together and not summed!

So manual say to use SUM, or to be precise hexadecimal SUM. But Should be binary ORed!
I lost a half of the day to discover that....

About calculator...
They are buggy!

python lut5.py -n4 '(~ i0)'
python lut5.py -n4 '(not i0)'

doesn't return same result! (the second one is correct)