Was it really? I'd like to see proof that it actually helped Python 3's adoption. Wouldn't new features from 3.5 like async/await have given people more incentives to switch had they been released earlier? (they would have for me)