re the question "has graph isomorphism been solved". this question asking about accuracy of an arxiv preprint claiming to prove graph isomorphism is now up to 9v. added an "answer" which gilles quickly deleted (even with one upvote in its brief appearance). think its great that this question stayed open without close votes and got good answers. however, its something of an anomaly. also, the 2nd top answer by Yuval now at 10v has an offhand remark
(If it had been, I would have known about it.)
which inspired my response which looks at how proofs in general are accepted by a scientific community incl wrt modern aspects (eg arXiv). agreed it is somewhat circuitous in answering the question but feel its all relevant.