Rich Shepard wrote:
I used natbib to manage the citations for my book. I'm now writing a
couple
of monographs that will be part of the take-home package from a water
quality
workshop in which I'll be lecturing. The references for the latter are
distinctly different from those of the former, and I'm considering how best
to accommodate them. Right now I have no comprehensive science library
database on my linux network, just books on shelves and journal
articles/agency reports in file cabinets and other shelves.
My initial thought is to build a new, separate natbib database for the
lectures. Almost certainly, each citation database will be specific to an
individual project, and entries in one would not be used for a different
project.
Regardless, I wanted to gain outside opinions from you fine folks who
have
solved this problem in many different ways. Your thoughts and
suggestions are
solicited. A while back I spent a little time seeking a bibliographic
database that ran on linux and that I liked. Didn't find anything then that
was worth the effort to build.
TIA,
Rich
My first reaction comes from a "database theory" perspective: avoid
redundant entries (same entry in two or more files). If you edit one
and not the others, things will get out of synch.
That said, I don't know that matters much whether you put everything in
one large (backed up!) BibTeX file, or use separate BibTeX files by
topic, so long as you can quickly determine which topic file would
contain a particular citation. I keep my references in five (so far)
BibTeX files, but their subject areas don't overlap, so it's unambiguous
where to go for a particular reference.
I've recently started using JabRef, and quite like it. (Previously I'd
just used a plain text editor to maintain BibTeX files.) Although I
haven't tried it, JabRef has a feature that will let you group
references by categories of your choice. So presumably you could group
yours by lecture.
/Paul
