IMHO the tone was quite annoying and overshadowed any good point the article could have about the complexity of proving concurrent algos. But slightly on a tan

IMHO the tone was quite annoying and overshadowed any good point the article could have about the complexity of proving concurrent algos. But slightly on a tangent, re: the QuestDB algo, wouldn't hazard pointers solve the concurrent map problem in a cleaner way and avoid a globally contended reader counter?