New arxiv interface for authors: go update your papers!

If you have any papers on the arxiv which have since been published, very likely you are amongst the many mathematicians who have neither added the journal reference to the arxiv article metadata, nor updated the arxiv copy to match the final (post-refereeing) version.

Go do this, right now.

The reason I’m so boldly telling you to do this is that it’s now easier than ever. On March 5, the arxiv rolled out a new submissions interface, and also added a new interface for managing your papers. This shows a list of your papers (and hints on how to claim others that a coauthor submitted, but that the arxiv hasn’t connected to you), with links for replacing or providing a journal reference. Adding a journal reference is particularly easy — it doesn’t require replacing the paper and the whole process involves filling out a single form. You can also give a DOI (which you can lookup on mathscinet, or your journal’s webpage), which will make it much easier in the future for automated tools to identify an arxiv eprint with the “official version”.

EDIT: (Ben) I would also suggest making sure all your papers are connected to your “public author identifier” especially if you have the same last name as another mathematician. Hopefully this will essentially make life easier for all of us.

8 thoughts on “New arxiv interface for authors: go update your papers!

  1. Of course a better arxiv interface is good. Also, one should certainly keep arxix postings up-to-date. The main reason for this is that
    the published version of the paper *doesn’t exist* in a practical sense: everyone that I know reads papers from the arxiv, not from journals.
    Also, in principle, an up-dated arxiv version can actually be better to read than the published version since the author has full control over it. For example, in my first paper the publisher chopped off the head of every arrow in every commutative diagram in the whole paper, making it completely unreadable. The arxiv version, on the other hand, is just right.

  2. I think the theory is that a DOI is a long term identifier; in two years, Springer (or JStore, or Cambridge University Press) might have rearranged every single web address on their web site, making direct URLs useless to retrieve any of their documents. In principle, there is some kind of process to “resolve” the DOI so that it will then re-direct to the right place…

    (Actual librarians tend to be very careful about issues of long-term storage and all that).

    By the way, thanks for pointing out this new arXiv page; I haven’t updated my arXived papers as often as I should, and it seems this will really make it much easier.

  3. You can “resolve” a DOI simply by tacking it on to the end of, or visiting that page and typing it in the box. For example, nearly always when you click on an “Article” link from MathSciNet, you’re actually following a link to

    Emmanuel has it exactly right — it gives some hope of URLs that survive the vagaries of online publishers website rearrangements.

    But there are also other excellent uses. For example, at present it’s annoyingly difficult to match up articles on the arxiv with their “final” versions on publishers web pages and MathSciNet. It’s easy enough for a human, by difficult to automate. Knowing this matching let’s you do some interesting things — in particular I’d love a browser plugin that could let me know whenever an article I’m trying to read, but which is behind a paywall, also appears on the arxiv. (Of course, for most of us, using the internet via rich institutions, this isn’t a problem. This, in itself, is a huge problem, because we don’t readily realise how difficult it is for many mathematicians to access the literature!)

  4. First, Ben, you gave an MIT-specific link there.

    Unfortunately, MathSciNet doesn’t have a link to this article, so I’m stumped too. Usually the article link is a redirect via, and you can read off the DOI from that. (Also, as of quite recently, MathSciNet is putting the DOI in a field in the BibTeX they produce.)

    Poking around, it seems that the Journal of Algebraic Geometry doesn’t obtain DOIs for their articles. Shame! :-)

  5. First, Ben, you gave an MIT-specific link there.

    Shoot, and I even went to the trouble of linking to the site outside MathSciNet’s paywall (I just forgot I had to take the proxy bit out of my URL). Anyways, fixed now.

    Poking around, it seems that the Journal of Algebraic Geometry doesn’t obtain DOIs for their articles. Shame!

    If only you’d told me that 5 years ago!

    In general, I find it really annoying that MathSciNet uses DOIs to link to the articles, but doesn’t just list them on the damn page.

Comments are closed.