“Advance” access and DOIs: what’s the problem?

A DOI, this morning

A DOI, this morning


When I arrive at work, the first task for the day is “check feeds”. If I’m lucky, in the “journal TOCs” category, there will be an abstract that looks interesting, like this one on the left (click for larger version).

Sometimes, the title is a direct link to the article at the journal website. Often though, the link is a Digital Object Identifier or DOI. Frequently, when the article is labelled as “advance access” or “early”, clicking on the DOI link leads to a page like the one below on the right.

DOI #fail

DOI #fail

In the grand scheme of things I suppose this rates as “minor annoyance”; it means that I have to visit the journal website and search for the article in question. The question is: why does this happen? I’m not familiar with the practical details of setting up a DOI, but I assume that the journal submits article URLs to the DOI system for processing. So who do I blame – journals, for making URLs public before the DOI is ready, or the DOI system, for not processing new URLs quickly enough?

There’s also the issue of whether terms like “advance access” have any meaning in the era of instant, online publishing but that’s for another day.

3 thoughts on ““Advance” access and DOIs: what’s the problem?

  1. biochemistri.es (@biochemistries)

    Yes I’m getting this all the time too. I know it’s pure impatience but I wanted to check out the supplementary files for something on PeerJ earlier, and since the DOI hasn’t been processed I can’t. In a sense it’s a minor annoyance but in cases like this it means I’ll have to wait to look at the code that’s often one of the most important parts of the work, especially in bioinfo (in this case just out of curiosity though).

    I’d be interested to find out what was causing the hold-up but no-one’s commented here since you posted… I’d guess it’s either the DOI registration agency (RA), such as CrossRef, rather than the International DOI Foundation (IDF)… I’m no expert though! I made a little post on DOIs here http://biochemistri.es/post/69580411732/on-going-deeper-than-the-article

    From a bit of further reading of the (not all that detailed) factsheets on the site it could be one of these RAs holding it up while it processes the application before finalising it with the IDF http://www.doi.org/doi_handbook/8_Registration_Agencies.html The RA has to use the Java client to send DOIs over and I’d guess they do this in batches after some sort of quality control. http://www.doi.org/doi_handbook/9_OperatingProcedures.html#9.1.2

    The only other clue is that the error page says “The DOI has not been activated yet” which could be the job of the IDF, but there’s just nothing on the site clarifying what this means. They state their role is “maintain the DOI system technical infrastructure and data”, which sounds to me like the RA is causing the hold up and the IDF just sets out industry standards. The DOI system is a lot less solid than you’d expect (to my understanding, inconsistency from having various RAs), even though it works great most of the time – there are a few links in the post on my site above to blog posts regarding that if all this is of any interest :-)

  2. alf

    It used to be that CrossRef had a backlog of articles to process, which meant that the metadata for an article (and its reference list) had to be submitted several hours before publishing an article, which was a bit problematic as the metadata would become public as soon as the DOI was activated, even if the article hadn’t been published yet. These days the DOIs are being activated almost straight away, so the article metadata can be submitted at the same time as publishing and the DOI is ready for use soon after. There may occasionally still be a delay, so it’s generally best to use the article URL rather than the DOI, at least for a while.

    1. nsaunders Post author

      Interesting, thanks. I notice that the CiteULike link on Nature articles uses DOIs – so bookmarking early articles often breaks, whereas the CiteULike javascript in bookmarklets such as AddThis uses URLs – which always works.

Comments are closed.