Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83 (MC830713); site epistemi.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!mcvax!ukc!cstvax!epistemi!rda
From: rda@epistemi.UUCP (Robert Dale)
Newsgroups: net.wanted,net.text
Subject: Bibliography Generation
Message-ID: <113@epistemi.UUCP>
Date: Sat, 25-May-85 20:28:24 EDT
Article-I.D.: epistemi.113
Posted: Sat May 25 20:28:24 1985
Date-Received: Sat, 1-Jun-85 13:38:41 EDT
Organization: Epistemics, Edinburgh U., Scotland
Lines: 36
Xref: linus net.wanted:5687 net.text:360

I'm sure this problem must have been met before.

We're currently thinking about building a shared online bibliographical
database, for use in the areas of linguistics, psychology, ai and philosophy
(to give you an idea of the kind of stuff we want to deal with), on a UNIXtm
machine.  We need some means of constructing, maintaining and using this
database.  We'd like to generate bibliographies directly from text files
containing citations.

Refer doesn't do quite what we want:  the form of citations produced in the
text isn't as flexible as we'd like, and the types of fields within records
don't really match up to our needs either.  The second isn't too much of a
problem, since redefinition is relatively straightforward:  we're thinking
about taking the Scribe set as our starting point, so that if we get Scribe
in the future, we can convert the database into Scribe format.  It looks
like we have to hack refer to solve the first problem.

Has anyone out there hacked refer to do things differently?  Has anyone gone
down this path already, and know of pitfalls etc.?  Has anyone laboured
night and day to produce new nroff bibliography macros?  Does anyone have an
entirely different solution to the problem?

Any suggestions or thoughts related to the above would be gratefully
received.


Robert Dale (rda@epistemi)

Snail:	University of Edinburgh
	School of Epistemics
	2 Buccleuch Place
	Edinburgh
	EH8 9LW

Email:	...seismo!mcvax!ukc!cstvax!epistemi!rda