Slashdot Mirror
Citation Managers For Unix?
A studious Anonymous Coward asks: "So, I now mostly converted to Linux, however one key applications that is missing: a tool like 'EndNote' or 'Reference Manager', which is invaluable for collecting and managing bibilgraphic resources when doing scientific research and writting scientific papers. These packages will do things like maintain a searchable database of your references which you can populate off of one of the public citation databases (pubmed, etc.). Additionally, it interoperates with Word to format your bibliography and (re)number your references. Is there anything like this for Unix?"
When it comes to collection of citations, I believe you are basically out of luck if you rely on PubMed as your primary source. As others have pointed out, the situation is quite different in CS, math, and other (typically formula-heavy) fields. I have actually considered making a PubMed front-end that searches their database for you (they provide URLs delivering the result as XML) and then reformat for BibTex. Who is interested?
Regarding management of citations however, I don't believe you can get anything better than using (X)Emacs in conjunction with the right modes. For managing your citation file(s), you have the BibTeX mode which is marvellously simple. For writing your paper, you use AUCTeX mode and the RefTeX minor mode. With RefTeX, you can search the citation database without leaving your emacs session and choose among matching citations to get the one you want in at the point. There is also support for managing references to figures and tables. For example, you can list all labels (to equations, figures, and tables) in a buffer, indented after sectioning.
But the inquirer confuses me regarding one thing. It is listed as a feature that EndNote can renumber your citations. In Latex, who cares?! You get all those details for free.
Lars
__
Reality or nothing.
Even though you're joking, you raise a serious point. This is actually a pretty common misunderstanding many people have about those who oppose intellectual monopolies.
I am personally opposed to copyright as a concept. But that doesn't mean I want to legalize plagiarism! People should be able to freely copy artistic and scientific works, but it should still be illegal to pretend that somebody else's work is your own.
I can't tell you how many times people assume that since you are opposing the right to monopolize certain ideas (or certain expressions), you are thereby automatically in favour of the freedom to falsely claim authorship of intellectual works.
One very nice thing about BibTeX is that you can get BibTeX citations for papers directly from the ACM Digital Library (sometimes), from DBLP, or from Citeseer -- so you download the paper, view it, and if you like it, cite it! There are also many large repositories of BibTeX entries for all manner of papers available, especially in the DB and OO communities -- just do a google search for "inproceedings" and a favorite author if you don't believe me. :-)
Here are just a few of the problems for a computer-non-expert: the changes to capitalization it does automatically (what, I need to say '{A}merica' to not get 'america'?),
So this requires a bibliography style file change.
the difficulty to debug if an entry is in the wrong format
This is pretty easy for me. Since my Perl script adds entries direcrtly from Melvyl, a problem with an entry results in hacking the script, and that particular problem never arises again.
This is happening increasingly as people create GNOME and KDE GUI interfaces for searching and adding to the bibliography database files. The GUI interface does most of the work in formatting the entries and error checking.
And of course, the journals would have no idea how to handle LaTeX code.
Journal are idiotic in that sense. It is actually MUCH MUCH easier for them to print something already in a well described typesetting language, than it is to print something in DOC format.
EndNote is our primary example of why we still mostly use Windows instead of Linux. I would love to see Niles port EndNote or to have someone else develop a product as easy to use.
I've asked them to port to linux several times. First was about 5 years ago. They said they had no demand for such a product. Last year they said they were working on it, but they were more than a little perturbed by the lack of standardization in software for word processing. I mean, they didn't know if they should write for StarOffice, WordPerfect, Applix, Koffice, or FrameMaker.
I, however, will continue to use LaTeX. I have a few very simple reasons. The first is that the format doesn't change in ways that cause incompatibility every few years to lock in the customer base. I know with confidence I will always be able to read and print my files. TeX has been reasonably standard for over 20 years now, since before Bill Gates was writing a BASIC interpreter for QDOS. The third reason is that my bibliography database management can be done in the UNIX way. I have small files and programs that each do their own small job well. The fourth reason is that I can load the word processor on as many machines as I like for free. The fifth reason is that if something is broken, I can fix it. The sixth reason is that if I am armed only with a text editor, I can edit my files.
Now, not all these reasons apply to other word processors, but taken on the whole I think I will be using LaTeX for a long time.
BibTeX is just fine.
:)
I have a Perl script set up to grab all email from Medline and automatically strip out the citations, format them, and add them to my database file. Even spent some time hacking bst files - now THAT is a weird language.
Anyone interested in the Perl script can email me. It won't work "out of the box" unless the email comes from Melvyl ie: the UC system. However, the principle is pretty similar for Medline and Melvyl emails. I just prefer the console Melvyl interface