I'd like to suggest that the recent letters complaining about reviewers' anonymity are on the wrong track. What we need is more anonymity, not less: we need double-blind mandatory anonymity. The main argument proposed so far is the unfairness of not being able to confront the reviewers' criticisms. But you don't need to know who someone is to be able to argue against their ideas. Reviewers' reports are spelled out clearly and can be rebutted without getting into personalities.