News for hackers. Stuff that matters.
Re:Should scientific articles be available free online? by PaulHoule on Thursday July 21, 2011 @05:08PM (#2791304)
No question about it.

If taxpayers pay $50k-$100k to scientists to write a paper, the least the government can do is spend another $5-50 to make it available for free perpetually.

Re:Should scientific articles be available free online? by rflrob on Thursday July 21, 2011 @05:24PM (#2791368)
Both of your estimates actually seem fairly low.

In biology, a single PhD student typically costs on the order of $50k/year in terms of stipend and tuition support, and will typically be an author on fewer papers than the number of years they are in grad school, so I would guess the personnel costs alone are in the $100k range, with materials and reagents potentially being a similar amount.

The publication costs in PLoS journals are 20-50 times higher than your high end[1].

All that aside, I absolutely agree that the costs to society of not doing open access are much greater than the costs of doing it.

[1]http://www.plos.org/journals/pubfees.php

Re:Should scientific articles be available free online? by sorbus on Thursday July 21, 2011 @06:35PM (#2791607)
> The publication costs in PLoS journals are 20-50 times higher than your high end[1].

I would think that the cost of actually making it available (no vetting, just throwing it online) is probably much lower than the cost of publication[1]. Note that, from your link, the expenses in the price include "peer review, of journal production, and of online hosting and archiving".

[1] If it really does cost them even 50% of the prices they're charging to put a single paper online, then either most of it is being eaten up in bureaucracy or they're investing it and using the interest to pay for the hosting (which would make a lot of sense, actually).

Re:Should scientific articles be available free online? by _delirium on Friday July 22, 2011 @09:53AM (#2793487)
It's not too clear to me why PLoS journals cost so much to publish in. In computer science, the open-access journals are typically free to publish in, supported by relatively modest funds from a hosting institution. For example, this is probably the top machine-learning journal, and I don't think it even has a budget or employees, just some donated bandwidth: http://jmlr.csail.mit.edu/

Where's the $3000-per-article difference coming from? PLoS journals do seem to have a more "magazine-like" production, while JMLR uses the typical CS/math approach of LaTeX typesetting. Is that the entire cost difference?