I'm not sure the inria team used the correct optimization for posit addition. You do yourself a disservice by treating a posit like a float and bifurcate the c

I'm not sure the inria team used the correct optimization for posit addition. You do yourself a disservice by treating a posit like a float and bifurcate the cross (negative/positive) from the negative/negative and positive/positive branches; since posits are twos complements the posit adder should be smaller, not bigger.