It's a good article, but it's not making the (obvious?) jump: the conference model in computer science is broken. No other discipline does it this way, and it

It's a good article, but it's not making the (obvious?) jump: the conference model in computer science is broken. No other discipline does it this way, and it creates for ourselves this terrible program committee problem. The journal model has plenty of its own flaws, but it at least allows for iterative work. (That is, instead of rejecting a work outright, a journal can work with an author on the flaws, even if that requires doing substantial new work.) This is not to lionize the journal model, which has plenty of its own flaws -- and indeed, I personally think computer science should use its current laggard status as an opportunity to find a wholly new model, preferably one that is much more amenable to practitioner participation.[0] [0] https://www.usenix.org/conference/atc16/technical-sessions/presentation/cantrill https://www.usenix.org/conference/atc16/technical-sessions/p...