[openbiblio-dev] Fwd: Best (or least bad) bibliographic interchange format?

Peter Murray-Rust pm286 at cam.ac.uk
Wed Aug 31 12:13:03 UTC 2011

latest mail

---------- Forwarded message ----------
From: Bruce D'Arcus <bdarcus at gmail.com>
Date: Wed, Aug 31, 2011 at 12:48 PM
Subject: Re: Best (or least bad) bibliographic interchange format?
To: beyond-the-pdf at googlegroups.com
Cc: ptsefton at gmail.com

On Tue, Aug 30, 2011 at 10:25 PM, Carl Leubsdorf, Jr. <carl at solvitor.com>
> Peter Sefton wrote:
>> I forwarded this to Bruce D'Arcus who invented CSL - the Stylesheet
>> processing language used by Zotero and Mendeley.
> Thanks for passing that along.
>> > But this general approach (with focus probably on JSON) would, in my
>> > view, be better than the legacy formats.
> I can see the value of this approach.  But (and not to be too
> on this) how immediately useful are future formats for people who, today,
> have hundreds or thousands of references already compiled in legacy
>  I suppose one answer is to import them into Zotero or Mendeley and then
> export to CSL.

So stepping back, let's define requirements. I was making the
following assumption, not really knowing much about the project:

1) project is a journal publishing extension for WordPress
2) if you're interested in structured reference metadata in that
context, you probably want to use it to enhance output

So if that's right, a primary goal should be embedded those data in
(RDFa or microdata) HTML output, and adding a bit of UI magic to allow
users to benefit from it. It would make sense, then, that there be a
pretty tight relationship between a storage format and that embedded
HTML, with ideally lossless and trivial conversion.

A separate matter is citation and reference formatting; I was assuming:

3) you might want a flexible formatting system, so CSL + citeproc-js
provides that ready-made

You may or may not need this.

Also, citeproc-js does not yet support embedded HTML output, so that
would need to be added to get what I'm envisioning.

So what are your requirements?


> We also may take the approach, since we need to save to NLM-DTD, to simply
> store the citations in a subset of the NLM tag suite, and work on
> CSL-compatibility for a later version.
>> JISC has just funded some work on HTML5 in the UK which I will be
>> involved in - one of the things we're looking at is how to represent
>> citations in HTML - I think that microdata compatible with the
>> Zotero/Mendeley/CSL format will be worth investigating. We will be
>> looking at automated bibliography formatting using CSL, probably in
>> Javascript - maybe you could work with us on that?
> I'm definitely willing to work on this; I suspect there will be some
> interest at NLM in this work as well.
> In the meantime I will have a closer look at CSL.
> Thanks!

Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/openbiblio-dev/attachments/20110831/e0e99293/attachment.html>

More information about the openbiblio-dev mailing list