Test evenE()

[goldbe]

This test, we should should use the even e=6 to make signatures (ie in the proving function). You are using the normal e=3 to make signatures (in the proving function). Could you change it? I'm referring to the test here.

https://github.com/osagga/TumbleBitSetup/blob/master/TumbleBitSetup.Tests/ProtocolTests.cs#L275

Also once you get this working, you should repeat the test several times and it should fail each time.

[osagga]

For this test I face two problems, the first one is when I generate a new RSA key (p, q) under e=6, BouncyCastle checks manually if e is a "KnownOddPrime" (here) and given how their code is written I just go into an infinite loop (I think they should've made it throw an exception).
So to avoid that, I generate an RSA key (p, q) under a valid e (like e=3) and then try to use those P and Q with e=6 instead.
Now the second problem is that when I try to pass (p, q, e=6) to the proving function, an exception is thrown from the Mod-Inverse ("ArithmeticException: Numbers not relatively prime.") function when calculating "d" (private exponent (here)
So that's why I had to use e=3 for proving and e=6 for verifying, what do you think?

[goldbe]

I see. Ok, let's sideline this test for now.

I think in order to get the test to work you will need a custom proving function, and to be honest at the moment I don't know exactly how this proving function needs to look like.

[goldbe]

Note: your first fix was good, but the second issue is what I am not clear on.