Help with understanding your code for my own implementation

[archit-p]

Hi!

I decided to create an issue since I couldn't find any other way to contact you.

I'm currently writing my own implementation for bulletproofs in c++. I've implemented the protocol (without aggregation) to prove that
v E [0, 2^n-1]
However I just can't figure out the technique you've used to disprove the statement, that is what if the above doesn't hold true. How does the protocol say that it doesn't hold?

As far as i understand the n value is assumed to be the length of the vector aL and hence the inequality will always hold. Any inputs regarding this will be greatly appreciated.

[AdamISZ]

Hi,

The issue of how to test such a setup seemed a little difficult/thorny to me, and at that particular point in time I was just mostly interested in figuring out the mechanics of the proof, so I didn't bother too much to figure out the "right" way to test correct failure with invalid proofs, as I wanted to move on to other things.

I think at the end of my writeup document https://joinmarket.me/static/FromZK2BPs_v1.pdf you'll find links to implementations by Poelstra, by the Monero guys and by Bunz himself, there is since then also some code and I think also detailed explanatory blog posts by Oleg Andreev et al at Chain. Perhaps in some or all of those you can find good ideas about how to write proper test code for an implementation.

[archit-p]

Thanks for pointing me to your write-up and other links. I will definitely go through them.