Doron Zeilberger has decided that pushing his papers through the journal publication process is a waste of time because the editors and referees take a long time, disagree with DZ about how a journal paper should be written, and then come back with stupid and inane feedback that shows that they failed to understand the point of the paper.
All true. And the feedback can be even more stupid and inane in conference reviews, when one gets feedback at all. But I draw a different conclusion for my own papers. The referees have been picked to be, as best the editor can find, interested in my papers and knowledgeable about the subjects. If even that hand-picked audience doesn't care for something about the way I wrote my paper, how can I expect any of my other readers to like it better? And if the referees grossly misunderstand some points of the paper, that's valuable feedback too: it tells me which parts of my paper are confusing and need to be clarified.
So, though I'm very fond of arxiv, I differ with DZ in not wanting it to become a replacement for our conferences and journals. There's a lot it does better than them, in making our papers available to a wide audience at a low cost, but it isn't a substitute for peer review.
On reading Doron's responses, I am now in fear of ever having to review a paper he's written (unlikely to ever happen, thank goodness). It's hard to tell when he's joking and when not, but the vituperation he reserves for the reviewers seems totally out of proportion to the actual comments themselves.
I also doubt that he'd admit that "parts of my paper are confusing and need to be clarified."
I find it interesting that DZ does not have comments turned on for his opinions.
I'm not sure there's a lot of significance to that — it seems to be merely a consequence of the fact that he's publishing the opinions as web pages rather than through any kind of blogging system.
One thing I wonder is what electronic communication could do for anonymous refereeing, which is partly based on the idea that two-hop communications take over a week physically. One of the most annoying things in a review is when a reviewer skim reads the earlier bits and assumes (based on prevailing approaches in the field) incorrectly the broad thrust of what you're doing and then criticises the details based upon that flawed premise. (The obvious thing to do is then to rewrite it saying not just what you are doing but very clearly spelling out what you're not doing -- whilst not appearing patronising.) This early misconception through inexplicit writing means the input of a good reviewer is "wasted". Maybe if there was a simple way to set up both-ways anonymous emails that might help iron out broad misconceptions. (Obviously there'd need to be some way to prevent the writer of a poor paper "spamming" a referee.)
Anyway, just a vague thought,
cheers, dave tweed