The Clinical Journal of Sports Medicine uses the following format for Digital Object Identifiers:

DOI: 10.1097/JSM.0000000000000xxx

Where xxx is a simple incrementing count, currently in the 200s, based on order of article publication. The latest issue has 26 articles, and they publish bimonthly, for a total of under 200 articles per year. Theoretically, if they increased publication to daily, they could reach 10000 articles per year. Hence, their 16 digit format will be sufficient for at least a trillion years.

Simple question: why? The usability is terrible; it's very difficult to transcribe by hand. They could chop the number of digits in half without any realistic risk of running out.

  • 12
    How about "10.1002/1521-3773(20020715)41:14<2596::AID-ANIE2596>3.0.CO;2-4"? :) – Marat Talipov Oct 27 '15 at 16:32
  • 3
    Perhaps they are considering micro publications, each data point can have its own DOI. – Kasper van den Berg Oct 27 '15 at 20:37
  • 1
    This question does not seem to be on topic for Academia as defined within our help center; see especially what not to ask. Your question seems slightly rant-y ("This is really lame, am I right or am I right?") and not answerable (the only people who know for sure why they set things up this way are folks who work for that journal; why are you asking here?). Those kinds of questions typically don't work well for our site format. – D.W. Oct 27 '15 at 23:59
  • 2
    I'm voting to close this question as off-topic because it's a rant and the underlying question is only answerable by the staff of the journal concerned, not by the general academic community. – David Richerby Oct 28 '15 at 0:55
  • @DavidRicherby, D.W. it's not necessarily a rant (ever heard of curiosity?), and some members of the general academic community may know the reasons for the particular case or other similar cases. – Franck Dernoncourt Nov 21 '15 at 14:39

This "answer" is really speculation - if you truly want to know their motivations, you have to ask them.

But I think the simplest and most likely explanation is that they don't consciously believe they will need that many DOIs; they just had to pick some format, and arbitrarily chose a number of digits that seemed more than adequate (as well as perhaps allowing for DOIs to be assigned in some other way than just serially; e.g. someday the DOI could identify volume, year, page number, etc).

They probably didn't even think about people needing to transcribe them by hand - and to be fair, in this age of electronic distribution and copy/paste, it's a pretty rare use case.

  • 3
    This. The only thing DOI has to be is: unique and admissible. How to achieve this is completely irrelevant. Someone may use UNIX time in nanoseconds when the DOI was processed by their server. Why not a very long fixed-width integer. – yo' Oct 27 '15 at 20:59
  • 2
    I suppose it's probably a coincidence, that they've chosen just enough decimal digits to represent the same contiguous range of positive integers represented by an IEEE double. – Steve Jessop Oct 28 '15 at 0:41
  • 2
    ... or by Javascript. – Mark Plotnick Oct 28 '15 at 1:21
  • @yo' Why not a very long fixed-width integer: usability, not having to set any limit, etc. – Franck Dernoncourt Nov 21 '15 at 14:40

Not the answer you're looking for? Browse other questions tagged or ask your own question.