Today, I was pleased to release version 1.5 of kcite. Follwing in my tradition of being unable to get the Wordpress plugin release process to work correctly, shortly after I released 1.5.1 which is the same thing, but with the correct metadata.

I’m quite pleased with this release. There have been some underlying changes to the technology which I will describe in another post, but for now I want to focus on what I feel represents a substantial improvement on previous versions in terms of functionality. The previous release added support for client-side rendering, which made things look nicer and will add more functionality into the future. However, from an authoring perspective this did not provide much advantage.

For the 1.5 release, I wanted to add new forms of identifier. Kcite started with the ability to cite digitial object identifiers to papers, as this reference to one of my papers shows (Lord and Stevens 2010) As a bioinformatician, references to PubMed also seemed like a good idea pmid:21414991[], even if, in most cases a doi could be used also.

However, for this release, I wanted to expand into two new areas. Kcite has come from the kblog project where we have been trying to improve the formality of publication using a blog engine, so that it can become an recognised part of the scientific literature. Organisations like arxiv have done much the same thing (with far greater success!) with preprint. It makes sense then that kcite can now also link straight through to here also (n.d.) Likewise, we wanted to support the push to data citation, which we have achieved courtesy of datacite. It is now possible to reference data sets also (Li et al. 2011)

And, finally, I have tidied up the presentation. Every item now has a visible URI in the reference list. I have made it visible because these identifiers should be public and present, not just providing an underlying link structure, although they do this as well now. In text citations link through to the bibliography, but also provide an outlink, direct to the resource also. Before processing, the underlying text citation displays the URI which should aid machine interpretability. This link will also be correct, since it is used to gather the metadata for the citation, rather expecting authors to tack a URI to a citation that already appears correct to them.

As always, new forms of publication raise questions and this release is no exception. We have, for instance, found that kblogs are useful for publishing grey literature such as bio-ontologies. But at the moment, rather embarrasingly, there is no way to reference these articles (or indeed this article itself!) with KCite. And, second, there is an issue of provenance. For instance, the notice describing the 1.4 release of kcite is now formatted with 1.5. It’s changed since it’s original publication because I have upgraded Kcite, as in turn will the presentation on this article change with the next release. I am hoping to address both of these issues in the near future.


Li, J, G Zhang, D Lambert, and J Wang. 2011. “Genomic Data from the Emperor Penguin (Aptenodytes Forsteri).” GigaScience.

Lord, Phillip, and Robert Stevens. 2010. “Adding a Little Reality to Building Ontologies for Biology.” Edited by Iddo Friedberg. PLoS ONE 5 (9): e12258.