Inadequate modelling of bibliographic citations (ODF pt 1, publicreview 1)

[Alex Brown <alexb@griffinbrown.co.uk>]

Dear all,

[This is a repeat of a comment made last year which remains marked as unresolved on JIRA. I am re-submitting unresolved problems which I see as critical in a quest for closure in the public review phase, with updated text where possible]

The modelling of bibliographic citations in ODF is *totally* *inadequate* for real-world content, and is not fit for purpose, seemingly failing to take into account any prior work in this field.

For existing XML models which *are* adequate for bibliographic citations, the TC might like to consider the relevant parts of the NLM [1] and NCBI [2] models, DocBook [3] or the CrossRef [4] schemas, or even ISO 12083 [5].

Faults with the current model include (but are by no means limited to):

* The things within a citation are modelled as attributes not elements. Bibliographic text such as chapter titles may contain formatted text, mathematics, etc., and this cannot be included in the attribute content model of ODF.

* Another consequence of using attributes is that sub-structures within content cannot be distinguished. So, for example, to enable link resolution (via CrossRef e.g.) citations are needed where possible to distinguish at least the family name from other parts of a name. In ODF names, and sequences of names, are a single opaque string only.

* Yet another consequence of using attributes is that features cannot be repeated. What if something has two authors? The current (undocumented) OO.o approach (of having author separated by ; characters) effectively invents a mini-language within the attribute and is poor practice.

* Major and necessary features of citations such as "article title", "doi" and "issue number" are not envisaged by the current model, yet hardly-used types of citation ("booklet" and "journal") are.

* Some content users, such as parts as the European scientific community (chemists particularly), have a strong preference for creating multi-part citations. These are not modelled.

To become acceptable, ODF must either remove the existing inadequate model, or replace it with a model which is fit-for-purpose; preferably one based on existing actual or de-facto standard.

[1] http://dtd.nlm.nih.gov/publishing/
[2] http://dtd.nlm.nih.gov/book/
[3] http://www.docbook.org/
[4] http://doi.crossref.org/doc/TechDoc.html
[5] http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=20866

- Alex.

______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email 
______________________________________________________________________