which smoothbrain was responsible for this convention? clearly for the linear penalty, d is an extension penalty as it scales with g. yet d is considered gap open for affine penalty and e is extension. and it's not these authors, i see it everywhere.