[Feature request] URN:NBN support

niklasholm
DOI is primarily an American system, but in other countries other systems of unique identification exist, most notably national bibliography numbers (NBN), which is an URN namespace, in use by multiple European national libraries.

I request that Zotero add a URN:NBN field in addition to the DOI field. To resolve the URN:NBN identifiers this service could be used and added as a lookup engine (which also works with DOIs btw).
  • adamsmith
    We'll have to think about a way to incorporate the wide range of identifiers available in the long run. Once we include NBN, there's easily justification for at least a dozen other systems, probably more, so this isn't trivial to solve.

    But
    DOI is primarily an American system
    is manifestly incorrect. DOIs fulfill a different function than NBNs, but the DOI system is broadly international with the EU having its own registration agency for its publications, for example (and Airiti doing the same for traditional Chinese publications).
  • niklasholm
    It's developed by American publishers, but whatever. It's not universally adopted, a lot of non-American publishers/archives doesn't use it for one reason or another.

    It doesn't matter. The point is Zotero needs to support identifiers commonly used in other parts of the world since not everybody uses DOIs. NBN is probably one of the most commonly used alternatives (in libraries not using DOIs).
  • adamsmith
    edited October 13, 2016
    But they're just not equivalent. NBNs are closer to OCLC numbers (which is another identifier, albeit a less-well-constructed one, that may be worth supporting, or LCCNs (another one) in their function.
    DOIs point to digital objects. NBNs point to any bibliographic resource, many of which may just be physical representations in a library (though many are digital, too).

    As such, NBNs are less critical for a reference manager. I've not seen them in a citation or seen their use recommended in a style guide. Doesn't mean we shouldn't support them, since uniquely identifying bibliography records is clearly useful information, but the equivalence is false.
  • niklasholm
    I never said they are equivalent in that context. What's your point?
  • adamsmith
    Point is
    1) DOIs do play a unique role in their importance as not just PIDs but components of citations/references of digital materials, and thus it's perfectly reasonable to prioritize them for any reference manager
    2) Supporting other IDs/PIDs for items is worthwhile, but there are a ton of them and so it's hard to do well. There's an older discussion on this with some proposals on how to implement this. I disagree that NBNs stand out from the rest of available IDs in a way that makes their support particularly critical.
  • niklasholm
    1) NBNs fill the same role where DOIs aren't available.
    2) It isn't trivial but not that hard either if the software is designed properly. I'd like some links to those discussions if you don't mind.

    I interpret your answer as basically: DOIs are most commonly used, therefor most important to support. That's fine. But they are not unique in that role. I think that not caring enough to support other systems is pretty arrogant.
  • bwiernik
    I think you are reading far too much into adamsmith's statements. He is simply stating that DOI as a system (1) serves a different more limited purpose than other ID systems [which also are designed to include print resources and unique items], (2) has much wider adoption than other systems [including outside the US], and (3) is actually required by citation standards [unlike any other ID system].

    DOI is intended as a universal system for stable digital object location and is being adopted as such over time. No other identifier system has this specific purpose or has acheieved this level of penetration. As a result, DOI has a great deal of functionality that other ID systems don't have. For example, one can use just the DOI to generate citations: https://blog.datacite.org/citation-formatting-service-upgrade/

    No one is arguing that supporting NBN would be bad--the argument is just that NBN is one of literally hundreds of ID systems that could be supported. Besides NBN, there are also ArXiV IDs, HDL IDs, OCLC numbers, LCCNs, ProQuest/UMI numbers, Web of Science numbers, etc. I'm not aware of a great argument for why any one of these needs support more than any of the others. Adding an identifier requires at least creating a field in Zotero and ideally, also integration with any sort of API or central management systems that exist for the ID. That is a large undertaking for even one ID. Just with regard to the fields being included, there is also the problem of including fields that the large majority of users will never need.

    A summary of current thinking regarding identifier support is here:
    https://github.com/citation-style-language/zotero-bits/wiki/Zotero-types-whiteboard

    You can find the full discussion of ID sets in the issues of that GitHub repository.

  • niklasholm
    Fair enough, I'll check it out and make my case for NBNs there. Thank you for your responses.

This is an old discussion that has not been active in a long time. Before commenting here, you should strongly consider starting a new discussion instead. If you think the content of this discussion is still relevant, you can link to it from your new discussion.

Sign In or Register to comment.