Bibliography Blues

July 12, 2011

The JWST discussion is getting vairy interesting. But meanwhile life continues. There are students to meet and papers to write. I am trying to finish writing up something I have been fiddling with for yonks. Let me burden you with three niggles about writin’ papers these days. The first of these I already tried out on the Twittersphere, but got a three-way split, so I am trying again….

We are not amused

I am writing a single author paper. Do I go for “I” or “we” or the passive voice ? Nothing seems right…

“In this paper, I consider the effect of X on Y…” Sounds a bit arrogant.

“In this paper, we consider…” So, Andy, is that the royal we ? Or are you having a personality crisis ?

“The effect of X on Y is considered…” By who ? Sounds a tad dry and formal.

Gaagh. What do I do ? I am bound to be wrong.

MNRAS long author list bug

Like many of you, I use BibTeX, and when writing a MNRAS paper, use the standard Blackwell supplied  LateX stuff including mn2e.bst file and a \bibliographystyle{mn2e} command. Mostly this works fine, but if you have a paper with a very long author list, it throws up an error when you run BibTeX:

Warning-you’ve exceeded 250, the entry-string-size, for entry Crenshaw1996 while executing-line 1223 of file mn2e.bst . Please notify the bibstyle designer.

and then LaTeX compilation bombs. Previously I have just excluded that paper and then added another \bibitem by hand. The Blackwell folk know this is a problem, because the README says

Note that there is a known bug in the .bst file: for very long author lists BibTeX reports an error. We are aware of the problem and hope to correct this in a future release

Unfortunately the README has said the same thing since Feb 2001. However, as often in geekdom, the community rides to its own rescue. I found a fix  written by Michael Williams at MPE. His file has the same name as the official Blackwell version, but I renamed it when downloading so I would know when I am using the fixed version. Thank you Michael !

Ex Libris Xerxes

What do people use these days for collecting and sorting their private libraries of research papers ? I have been oscillating around various options and not quite stabilising. At two ends, the constraints are clear. One wants a .bib database for squirting things into our LaTeX documents; and one wants to grab things from ADS and arXiv . In ADS itself you can construct private libraries and export them, and I use this a fair bit, but it doesn’t give me the sort of hands-on metadata editing I sometimes want, or the sorting, grouping, tagging and general manipulative ability that you want. For that sort of thing, there are quite a few commercial packages, like EndNote, Papers, Bookends, etc, but I have always used the free Java app Jabref. This is v.good, but it doesn’t give me the ability to grab and add things from ADS.

Some publishers have tried to solve this problem via a web interface, and of course lock you into their world while they are at it. Springer tried CiteULike  (worst —  name — ever) and Nature publishing tried Connotea . Being server hosted, they can add a social networking side. Join a group, share your favourite papers with your chums etc. Then I discovered Zotero. This is a kind of hybrid. Its a firefox extension that feels like a proper app but works with ADS, giving a one-click import. It has FOLDERS not just bloody tags so you can organise things just as you want them. But it also has that social groupy thingy. Its lovely. But I wanted to use Chrome… or at least not be locked into Firefox. So I decided to give it up.

So then I discovered that the trendy young things these days use Mendeley. It is a commercial enterprise, but with an open feel. The basic version is free, but you pay for premium service like more storage and extra features. There is a desktop client and a server side, so they try to get the best of both worlds that way, but that does mean you have to keep synchronising. The most annoying thing has bee the lack of sub-folders (I like a hierarchy myself) but I heard a rumour the new version fixes this.

So anyway I will keep at it for a while. But right now my chain of action is (i) Browse ADS. (ii) Grab into Mendeley. (iii) Export folders to bibtex. (iv) Fiddle with bibtex with Jabref (v) Run LaTeX.

There is probably a discussion on Astrobetter  somewhere, but of course I can’t be arsed to go and look for it. Stufff to do.