preview of next version of kcite
Phillip Lord
It's not quite ready for a full release (although it's in our mercurial
if you want to try it), but I thought that people might be interested in
some HTML dumps from the latest version of kcite.
It now generates bibliographies on the fly, using javascript
(citeproc-js and a few other bits). The "even bigger" files show some
files from a post with (nothing but) 240 references in. The "reality"
files are a post of the first part of a real paper. Currently, the in
text citations work but the bibliography is broken for all but chicago
format -- working on it. Switching between these requires changing code
at the moment, but as rendering is all client side, it just needs a
button or two to allow the reader to select their style.
I've found this to have reasonable performance -- with 240 references,
it makes my CPU churn a bit, but the browser remains active. Let me know
if this is not true for you. With 20-30 references (which is, I think,
more realistic), rendering is too quick to see.
You can't tell from this (as they are static dumps), but kcite caches
better now, so there is one external lookup per week per reference.
Full release is a few weeks off, but thought people might be interested.
http://homepages.cs.ncl.ac.uk/phillip.lord/test/
Next release is going to do cito better. After that it's ajax calls back
to the server, and then configurable caching. All of this will happen at
some undefined point in time, since I am not a neutrino and am bound by
the speed of light.
Phil
--
Phillip Lord, Phone: +44 (0) 191 222 7827
Lecturer in Bioinformatics, Email: philli...@newcastle.ac.uk
School of Computing Science, http://homepages.cs.ncl.ac.uk/phillip.lord
Room 914 Claremont Tower, skype: russet_apples
Newcastle University, msn: m...@russet.org.uk
NE1 7RU twitter: phillord
Carl Boettiger
Phillip Lord
Carl Boettiger <cboe...@gmail.com> writes:
> Phillip,
>
> Thanks for this preview, the examples look great. Parenthetical references
> (i.e. Chicago style) are a very nice touch.
Once I release this, it should be configurable by the blog owner. Next
step is to have it configurable by the reader.
> Minor question: Will there be a way to include references manually in the
> list that don't have a pmid or a doi?
At the moment, there isn't. It's actually very easy to add either using
embedded JSON or an (overly long) shortcode. It is next on my list of
things to do. However, I don't want to delay the release, because there
are quite a few things which are "easy to do".
It's next on my list, as it is causing me pain and grief.
> Currently these have to be added as separate footnotes, which is a bit
> annoying when you have one old jstor paper that doesn't have an id (or
> a general url you want to cite).
My current list of identifiers that I want to support goes something
like: kURLs (that is a URL to a knowledgeblog article which declares
it's own metadata), other URLs where possible, arXiv (with added
trackbacks!), ISBNs, ISSNs.
As always, this is on an ad hoc basis, and reflects my own bias. It will
also reflect how easy it is. Code contributions are always welcome:-)
> Also looking forward to the cito integration, even if it's only at
> light speed.
Also easy, although as yet, I don't know how people are going to use
this; CiTO doesn't fit into my tool chain.
Phil