The mathematicians are revolting! Scientific presses charge university libraries exorbitant fees for subscriptions to peer-reviwed journals. The journals published by the nonprofit Association of Computing Machinery, the American Mathematical Society, and the Society for Industrial and Applied Mathematics cost 20-30 cents per page. By contrast, Elsevier, Academic Press, Kluwer, and other commercial publishers will often charge over a dollar a page. Thats quite a nice profit margin. And the charges faced by university libraries are so severe that many are cutting back substantially on their subscriptions.

But there is a limit to what univeristy libraries can do. Elsevier, for example, bundles up hundreds of its journals in the form of Elsevier Science Direct. Universities cannot possibly refuse a subscription to Science Direct -- they might feel OK with not subscribing to this or that prominent journal but they cannot afford to leave out so many of them. As a result, Elsevier manages to charge a lot more for Science Direct than it would earn if it sold each journal in the package individually.

What is so frustrating about this is that, really, the community of mathematicians does not need commercial publishers. We don't get paid for the editorial work we do for the journals. The crucial element of these journals is peer-review -- and reviewers don't get paid. There is no reason why the math community cannot simply establish a host of online peer-reviewed journals which would be available to everyone free of charge.

However, online journals have not taken off. Sure, there's a bunch of them around; but they don't seem to garner as much respect as is accorded to the traditional paper-based journals. They're not prestigious. You would not submit your best paper to an online journal. Why not? I haven't the slightest idea. And it doesn't help if the online journal has an impressive editorial board.

Online journals are not prestigious because everyone believes that online journals are not prestigious and, as a consequence, no one submits their best papers to them. Its purely a state of mind -- and one that doesn't seem to be going away.

The whole thing seemed hopeless until recently when the entire editorial board of the Journal of Algorithms resigned over the high cost of the journal and unanimously agreed to start a new journal, the Transactions on Algorithms, to be published by the Association for Computing Machinery, a non-profit society.

(Actually, this is not the first time something like this has happenned -- part of the editorial board of a machine learning journal resigned a couple of years ago -- as well as the entire board of a logic journal. But the Journal of Algorithms is pretty prestigious so this is a milestone).

I hope this is the first step to a revolt of the mathematicians.