LATEX-L Archives

Mailing list for the LaTeX3 project


Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
"P.W.Daly, MPAe, Lindau, Germany" <[log in to unmask]>
Reply To:
Mailing list for the LaTeX3 project <[log in to unmask]>
Tue, 4 Nov 1997 09:39:52 +0200
text/plain (84 lines)
In the discussion on standardizing a journal class, my natbib package has
been frequently mentioned (positively, thank heavens), mainly by
Sebastian and Phillip. I would like give a brief overview of it for those
who are unfamiliar to it. Sebastian did ask me some time ago to write a
TUGboat article on it, and, well, that is still on my todo list.

Natbib.sty was originally made up to handle author-year citations. In
this sense, it was far from the first such package. There were apalike,
newapa, chicago, harvard, named, and so on. Each of these had its own bst
files, which could only be used with that package. New \cite commands
were necessary in every case, if only to distinguish between citations
in text like Jones et al. [1995]. and parenthetical ones
[Jones, et al. 1995]. That is the basic requirement for author-year
citation (which apalike does not even fulfill). Additional requirements
are commands for authors only, year only, both citation variants with
full author list [Jones, Baker, Williams, and Gouge, 1995]. Chicago and
especially harvard can handle these. But: each one has a different set of
\cite variants to this. In that sense, natbib is no better, for it uses
yet another set.

However, natbib has developed further to have two main advantages over
chicago and harvard: 1: it can read bst files intended for the other
packages; 2: it can switch between numerical and author-year citation

Under point 1: natbib can even read plain.bst and the other numerical
styles, but then can only produce numerical citations. I provided
plainnat.bst, unsrtnat.bst, and abbrvnat.bst as natbib replacements for 3
of the 4 standard bst files that can be used with either numerical or
author-year citations.

Point 2 means that one can write
  the argument of Bright [1984] has been refuted [Smith et al. 1995]
for author-year citation mode, and if one changes to numerical citations
by selecting a \usepackage option, one gets, without any further change
in the source text,
  the argument of Bright [23] has been refuted [54]
(This requires that one uses an author-year bst, but any one will do,
even those intended for harvard or chicago.)

It is safe to say that natbib can do anything the other packages can do,
and more. Some people might object, saying harvard can give citations
with full author lists on the first citations, and short lists
afterwards, whereas natbib can do this, but not automatically. I can
refute this simply by saying that version 6.7, which is about to be
released, also has this automatic feature.

Phillip has talked about the natbib native format for \bibitem, which
looks like this:
 \bibitem[Jones et al.(1995)Jones, Baker, Williams, and Gouge]{key}...
This can only be used with natbib.sty; none of the others will treat
this right. However, if natbib were to become quasi-standard, this is no
problem. The only thing that worries me is that if I were starting from
scratch today, I would invent \natbibitem with three clearly separated
arguments for the year, short and full lists. As it is, I rely on the
parentheses to act as delimiters to separate the three parts of the
citation information, something that could cause trouble if an author's
name contains parentheses. I have talked to Oren Patashnik about this,
and there is a tendency to put square brackets into first names to
indicate how the name abbreviated. I am toying with the idea of switching
to \natbibitem once BibTeX 1.00 comes out. But of course, older syntaxes
would still have to be supported.

My other major contribution to bibliographying is   custom-bib  or
makebst. This allows one to make up bst files from a gigantic master by
selectively taking lines of code, governed by a set of options. An
interactive menu system helps select these options, putting them into a
docstrip batch job file, for future editing/correcting or rerunning on
updates to the master file. There are over 100 options available, with
the number growing.

This will do for an overview. Now I should concentrate on writing the
TUGboat articles, or getting the latest versions of natbib /makebst

   Dr. Patrick W. Daly                    Tel. [+49] 5556-979-279
   Max-Planck-Institut fuer Aeronomie     Fax. [+49] 5556-979-240
   Max-Planck-Str. 2
   D-37191 Katlenburg-Lindau              Internet: [log in to unmask]