After the post last week and the crazy discussion that followed I would understand that you feel you have heard enough about citations in markdown. But I had the feeling last week that something was still missing, and I have done some more thinking. What we have so far:
And I looked at how Wikipedia is doing this, and they use a) links, b) citations and c) footnotes. If Wikipedia thinks that it can’t do without citations and do everything as links, then maybe we also shouldn’t enforce this for scholarly texts.
I think what we need is the best of both worlds. We should use the Pandoc citation workflow, as it is similar to what we are used to from other authoring environments, and we get good citation style support, including more complex formatting of references. Some reference managers already support copy/paste of Pandoc citation keys. The inclusion of a bibtex file with a scholarly markdown text is also a bonus, as it allows the automated extraction of citations, e.g. by manuscript submission systems.
We also want to support a simpler solution for shorter texts or when people don’t want to use a separate bibtex file. Here we would add the citations as links, ideally in a syntax very similar to Pandoc citation keys:
Johnson [@Johnson2006] didn't agree with ... [@Johnson2006]: http://dx.doi.org/10.1002/aris.201 "Data sharing in the sciences"
We need to write a tool that parses the markdown before Pandoc, fetches the citation metadata for these links in bibtex format (e.g. using CrossRef Content Negotiation), and adds them to the existing bibtex file (or creates a new bibtex file). The next time the markdown is parsed, the citation is already “cached” in the bibtex file. Those people who don’t have such a tool would see the citation as link (???), with the essential information (DOI or URL) preserved so that a downstream tool can fetch the bibliographic information. Some people were worried about typos in DOIs and URLs. They can add additional information - e.g. the title of the paper - in double quotes to allow checking of the correct DOI.
This workflow now makes a lot of sense to me, as it uses existing solutions, but also allows for easy entering of citation information in a way similar to the knitcitations and kcite tools. As I use jekyll and am a Ruby developer, I will implement the citation parsing as a jekyll plugin.
Using YAML Frontmatter with CSV
CSV (comma-separated values) is a popular file format for data. It is popular because it is very simple: CSV is text-based and any application that can open text files can read or write CSV. This makes it a good fit for digital preservation. ...
Upcoming Changes to DOI Content Negotiation
DOI content negotiation is one of the oldest DataCite services, launched in 2012. Content negotiation makes it easy to fetch DataCite metadata in other metadata formats, for example BibTeX or schema.org, or as formatted citation in one of more than 5,000 citation styles. ...
2020 Strategic Priorities for Services and Infrastructure
In a blog post four weeks ago DataCite Executive Director Matt Buys talked about the DataCite strategic priorities for 2020 (Buys, 2020). In this post we want to talk a bit more about the strategic priorities for this year we have regarding services and infrastructure work: a) ...