Overlays over arXiv
Much has been said about overlay journals [@url:gowers.wordpress.com/2013/01/16/why-ive-also-joined-the-good-guys/] The idea is simple; the journal essentially becomes a selector, a channel, with the paper itself being hosted elsewhere, such as arXiv.
This holds a certain amount of attraction for me; I already post my new papers on arXiv. I have been posting them here also [@url:www.russet.org.uk/blog/1713] This works well, but is hampered by technology. Mostly I write papers in LaTeX, and I have written tools to make these suitable for Wordpress [@url:www.russet.org.uk/blog/1740] these work well enough to publish an entire thesis [@url:themindwobbles.wordpress.com/2013/01/02/phd-thesis-table-of-contents/] However, the process of doing this is not slick [@url:themindwobbles.wordpress.com/2012/06/14/converting-a-latex-thesis-to-multiple-wordpress-posts/] For instance, when trying to publish one of my own papers, I have had problems as I used a theorem environment [@doi:10.1186/2041-1480-1-S1-S4] While PlasTeX is a nice tool, the key problem is that it is fundamentally a different interpreter from TeX. Eventually, perhaps, LuaTeX will get an HTML backend, but until this happens the system will always fail in some cases.
So, I wanted to investigate whether it was possible to build Overlay functionality into a personal publication framework, such as the Wordpress installation I host these articles on. Well, it turns out combined with the tools that I have written for manipulating metadata [@url:knowledgeblog.org/kblog-metadata] it is relatively simple to do so; my first attempt at this is now available for my OWLED 2013 paper [@url:www.russet.org.uk/blog/2366] The title, authors (just me in this case), date, abstract and PDF link all come directly from arXiv. Full text is not available from arXiv --- anyway it would suffer from all the issues described earlier; in the end, the PDF is probably the best representation of this paper. I have supplemented this with a plain English summary, something that I have wanted to do for years, but have not managed to start. If the reviewers will allow me to do so, I will also attach these when they become available.
The code for this is not quite ready to release yet: however, it will potentially work over any eprints repository, and I have connected it up to Greycite also [@url:greycite.knowledgeblog.org] so it can be used over any source that greycite can interpret.
All a little clunky, but I think that this is the future. The Journal is dead, Long Live the article.