allow both DOI and arxiv id

[adasilva]

Many papers have both a published version (DOI) and an arxiv version which are extremely similar, if not the same. Many of the published papers are behind paywalls, while the arxiv papers are free. Why not allow both a journal DOI and an arxiv id in the same gist?

[jure]

Interesting, I never looked at it like that. Can you provide some examples to get a better idea/paint a clearer picture? It certainly feels like that is the case and we should allow both identifiers on a single Paper/Gist.

Would you be willing to come up with a pull request that adds that feature?

[adasilva]

Example (a paper I wrote about a year ago):
http://arxiv.org/abs/1307.6550 (arXiv ID 1307.6550)
http://journals.aps.org/prb/abstract/10.1103/PhysRevB.88.195411 (DOI 10.1103/PhysRevB.88.195411)

I am just learning Ruby on Rails. Adding this feature might make a good project for me to learn on, but since I'm new to the language it might take a little while.

[jure]

Cool, thanks for the example. These two papers do in fact appear to be nearly identical.

What percentage of papers on arXiv would you say have a published variant somewhere else? Is it possible to connect the arXiv paper with its sibling without human interaction?

Take your time and I'll be happy to review the PR.

[adasilva]

I'm from a physics background, and in my field close to 100% of peer reviewed publications are also on arxiv - it's very ubiquitous to submit to a journal and arxiv at the same time.

Although still a large percentage, it's not quite so high the other way around because (1) there's some lag between submitting to arxiv, getting accepted for publication, then to actual publication; (2) some papers posted on arxiv will get rejected and never be published; and (3) a small number of papers on arxiv are not meant for peer-review publication (e.g. notes on a specialized topic).

Getting the journal ID from arxiv should be easy: arxiv has a place for authors to update the journal link once it's available. Getting the arxiv ID from the journal DOI would not be so easy, but google scholar is able to do a pretty good job of matching journal to arxiv most of the time, so it is possible.