Sorry I part company with you Anthony. There is little reason to redeposit an OA item just in case the firm goes bust or is sold. It is simply wasted effort and almost every researcher will see it as such. The URL/DOI/handle is far superior as it guarantees correctness and no hacking or deception. Imagine posting a revised version of an article to a repository (unrefereed) in order to deceive. BTW, the cost of maintaining a redirection is minimal for a publisher.
One might as well manufacture citations to refer to your preferred version. Now if you mounted an argument on archival preservation through multiple copies, that at least has some legs. Not a lot, but the argument is valid. I discard the argument about OA checking because that assumes that you need to know this and you demand to know it simply - a totally second order issue of interest only to administrators, which a competent computer scientist can solve. All this is more important if the OA version is not a pdf, as it could be and sometimes is (HTML, XML, etc). The Australian research councils agree with me. Arthur Sale -----Original Message----- From: goal-boun...@eprints.org [mailto:goal-boun...@eprints.org] On Behalf Of Andrew A. Adams Sent: Wednesday, 27 November 2013 11:41 AM To: Global Open Access List (Successor of AmSci); Jan Velterop Cc: Rick Anderson; scholc...@ala.org; LibLicense-L Discussion Forum Subject: [GOAL] Re: [SCHOLCOMM] Unanimity (Re: Monographs) Jan, There are reasons for requiring green deposit even where an article is already OA on the journal. First, that canchange - one of my articles was published in a new Journal (Vol 2(2)) which a couple of years later was sold by bepress to iley and became closed - very annoyingly to me on two grounds - I was not informed of the change and the original URL stopped working. Second is the very practical measure that it is easy for an institution to check whether every article deposited has a full text paper in the repository (simply by checking that a suitable document is there - the fact that it would be possible for an author to upload a blank document requires an assumption of malfeasance far beyond the likelihood of it ever happening - the chances that sooner or later someone would spot and report such an upload are so high that very few would be foolish enough to do this). Compare this to checking (regularly because of my first point) that the URL (which is more likely to a web page rather than adeep link tothedocument) gives open access, which takes a ridiculous amount of work. Institutions almost universally already collect for very good reasons the meta-data of their researchers' output. Adding the requirement to submit the full text of the accepted version is a very small amount of extra work done in a scalable manner. Everything else does not properly scale. On the point about libre OA and gratis OA, I'm afraid you are wrong about open in English meaning the same as libre. Open has exactly the same problems as "free" in terms of being overloaded. I'm working on a paper at the moment on this issue, but the simple pointer to this is the use of the word "open" in the two phrases: Open Educational Resources (OERs), in which open generally means "libre" Massively Open Online Courses (MOOCs), in which open generally means "gratis" None of the main MOOC platforms have libre licensing of the content, not even the non-profit ones. MIT's OpenCourseWare (not the first or the only OER resource but a major instance of it) specifically provides for a CC-BY-NC license. -- Professor Andrew A Adams a...@meiji.ac.jp Professor at Graduate School of Business Administration, and Deputy Director of the Centre for Business Information Ethics Meiji University, Tokyo, Japan http://www.a-cubed.info/ _______________________________________________ GOAL mailing list GOAL@eprints.org http://mailman.ecs.soton.ac.uk/mailman/listinfo/goal _______________________________________________ GOAL mailing list GOAL@eprints.org http://mailman.ecs.soton.ac.uk/mailman/listinfo/goal
