Re: [OpenAccess] DOAI vs. DOI resolver

Hi all, A question elsewhere about doai.io reminded me of this, and that no-one ever replied. So, a reply :-) It would certainly be nice to support green OA (which, historically, we've not always paid very much attention to...). I think *switching* to replace dx.doi.org might be a bad idea. For example, it's a hassle for anyone who does have access and wishes to see the original copy rather than a MS version. Before switching, we'd also need a better understanding of how often it updates/confirms archived copies are still available, or whether the original has become accessible (consider a PNAS paper; they have a moving wall, so in the first six months you'd want a self-archived copy, and after that would prefer the journal.) But we can still use it. Some options: a) We display a second "free copy, if available" link after the DOI, for all DOIs, trusting that it will fail safely - it probably will; see, eg, http://doai.io/10.1093/femsle/fnw043 b) We process our list of DOIs which exist on-wiki, look them up through doai.io & dx.doi.org, flag all the ones where the two differ; then add a "free copy available" link to these citations in particular. Run every few months as needed. The first raises false optimism; the second might involve a lot of update editing. But they're workable. Thoughts? A. On 28 February 2016 at 14:36, Federico Leva (Nemo) <nemowiki(a)gmail.com> wrote: -- - Andrew Gray andrew.gray(a)dunelm.org.uk