Mathematicians aim to take publishers out of publishing
January 18, 2013
Mathematicians plan to launch a series of free open-access journals that will host their peer-reviewed articles on the preprint server arXiv, Nature News reports. The project was publicly revealed yesterday in a blog post by Tim Gowers, a Fields Medal winner and mathematician at the University of Cambridge, UK.
The initiative, called the Episciences Project, hopes to show that researchers can organize the peer review and publication of their work at minimal cost, without involving commercial publishers. …
Many mathematicians — and researchers in other fields — claim that they already do most of the work involved in publishing their research. At no cost, they type up and format their own papers, post them to online servers, join journal editorial boards and review the work of their peers. By creating journals that publish links to peer-reviewed work on servers such as arXiv, Demailly says, the community could run its own publishing system. The extra expense involved would be the cost of maintaining websites and computer equipment, he says.
That cost is not small, but it could eventually be provided in part by the journals’ users. The arXiv server, for example, costs about US$826,000 a year to run, and is funded by the Cornell University Library in Ithaca, New York; the Simons Foundation in New York and institutional members.
Demailly says that he first thought of open-access electronic journals that overlay arXiv eight years ago, but the concept became a reality only last June, when he was contacted by the Centre for Direct Scientific Communication (CCSD), based in Villeurbanne, France. The CCSD, a unit of the French National Centre for Scientific Research, develops open-access repositories such as the multidisciplinary archive HAL, which mirrors the arXiv site.
For the Episciences Project, the CCSD plans to create a publishing platform that will support online peer-reviewed journals. Each journal, or ‘epijournal’, would have its own editor and editorial board, and authors could submit their arXiv-posted papers to their journal of choice. The journal would then organize peer review, perhaps using workflow software provided by the CCSD.
Peer-reviewed papers would be posted on arXiv alongside their un-reviewed versions. A central committee (led by Demailly) would manage new journal candidates and make recommendations on paper formatting, but each journal would be free to set its own policies (including whether to charge for publication).
(More)
Comments (9)
by Peter the printer
‘its own editor and editorial board’ precisely what publishers provide!
by Peter the printer
What arrogance, so, lacking understanding on what goes into publishing and what skills are needed, they think they can do it all themselves. The result will be a DIY mess with mistakes on every line which will be typified as ‘typos’ but will in fact be lack of English skills since mathematicians are too important, and busy, to learn English. Amazing how so many people think that others’ skills are easy or non-existant, but theirs are unavailable to us non-mathematical lay persons. As the internet has spread, English has gone downhill due to the ignorant thinking they have just as much right to express opinion as anyone [perhaps they're right, but they should try to be professional about it]. Next, mathematicians make movies, record CDs, build houses and invent new engineering applications; no need for skilled professionals after all.
by DAG
I rent 0.1Tb (1Tb bandwidth allowed per month) for $120 / year… this is from a top notch facility with many bells and whistles and 24hr support.
So my mind is blown by this nearly $1million per year they’re paying to keep their system running…
by MikeB
The server that hosts arXiv costs $826,000 a year to run? To host _just_ arXiv? That doesn’t sound right at all. Anyone know of other options? Could you do it cheaper by renting space from Google or Amazon?
by Whittaker
$82600 per year is actually okay, consider that they hold (shown on their homepage) 814780 papers. That is like a little more than 1 dollar per paper per year.
Much less than the fee demanded by “intellectual property mafia”.
by MikeB
The summary said $826,000 but does it cost $1 per paper per year to serve those papers? Perhaps an order of magnitude less $.10US/paper/yr ? Again, it seems more likely that $1million/yr server is actually handling much more than arXiv data, so what portion of that cost should be attributed to arXiv data?
by Amit
You are right, MikeB. I have to provide server space to anyone who uses my documentation platform. It would be unfeasible if it cost $1 per document to store.
That said, one reason why it costs so much to maintain the arXiv server is that they offer downloads in multiple formats. Unless we have a universal convertor (like a transcompiler) that generates content in the desired format on demand, this will continue to be a problem.
by t
clearly you are not taking into account that arXiv is a dynamic resource with many services which require curation, moderation, user assistance, software development, cooperation with related service, long term archival, etc.
your argument that hosting 10^6 static pdf files shouldn’t be much more than bandwidth charges does not even come close to grasping what operating arXiv and all it’s value added services entails.
so it’s probably the correct ball park figure.
Cheers
T.
by Amit
T,
Firstly, you are over-estimating the cost structure for running arXiv. I should know, because I design and implement platforms for manipulating all sorts of data.
Secondly, even if you were right about it, that does not take away from the fact that there is a much more efficient way to manage this data which is possible.