@eugene, can you exapand? I'm wondering the same thing. cc @Wintermute

Test jumpBackOnNegative for java :
sstttnssnsnsssttntssntnstssstntssssnsntt

should result in exception.
sstttn // push
ssn // Exception
..

✘ Exception not expected: java.lang.NullPointerException

@eugene -- how? in both the interpretations F2's arity is one.

it depends on F2's arity

Ok.. that worked. Thanks!
I guess it would be helpful to specify that in the kata details.

It's using Knuth's floor mod. I believe I had tested every statement via Haskell, but if that's not correct via Haskell, I'll just say that this variant uses Knuth's floormod.

The testcase I am referring to is:
test[floorMod4] ArithmeticEdgeTest

java

Which language are you working with?

In the final test case : ssttstnsssttntstttnst

The modulo -5%3 should evaluate to 2, but the test expects 1. Please let me know where I am wrong.

ss ttst n // push -5
ss stt n // push 3
tstt // evaluate b%a
tnst // output.

✘ Should be the remainder with the sign of the divisor expected:<[1]> but was:<[2]>

What should be the outcome in such a case :

F1 x F2 y + 1

is it F1(x,F2(y+1)) or is it F1(x,F2(y))+1 ?

The grammer isn't clear about this.

No, so I guess there is a bug (at least)

Will pass2() reduce expressions of the form 2x3 ?