The current system of journal publication has odd incentives -- paying to be published and doing peer review for free -- and should be revised in some way. Unfo

The current system of journal publication has odd incentives -- paying to be published and doing peer review for free -- and should be revised in some way. Unfortunately though the journals provide the metric by which basically all researchers are judged. I agree that the world would be a better place if research were freely available, but the way Sci-Hub is doing this does not help the process, unless it stresses the system so bad financially that it breaks and is replaced by something else. I have wondered about the possibility of all articles being hosted freely on sites like arxiv.org with some kind of web of trust rating system: anyone could vote up an article but people in specific fields could choose to view the aggregate rating of respected researchers in their field. In any case, until a better rating system for researchers emerges, we can't get rid of the journal publishers.