Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!olivea!olivey!jerry
From: jerry@olivey.olivetti.com (Jerry Aguirre)
Newsgroups: news.software.b
Subject: Re: How do I get out of this mess?
Summary: program to remove duplicate articles after expire -r
Message-ID: <50065@olivea.atc.olivetti.com>
Date: 3 Jan 91 22:14:39 GMT
References: <1990Dec29.040728.8886@intacc.uucp>
Sender: news@olivea.atc.olivetti.com
Organization: Olivetti ATC; Cupertino, CA
Lines: 21

In article <1990Dec29.040728.8886@intacc.uucp> mann@intacc.UUCP (Jeff Mann) writes:
>Now I have a slew of duplicate articles, and I am wondering how to get
>rid of them. What will happen if I try to rebuild the history file
>with mkhistory - will the duplicate message id's screw up the database?

The B news expire will add all the duplicates to the history text file.
It does not match up the hard links correctly but they are all listed
and will eventually get expired.  I have no idea what it does with the
DBM/DBZ file.

I wrote a program to extract the pathnames from the history file of all
but the first entry in each news group for each message ID.  Thus if
article <123@site.com> exists in news.admin as 100, 109, and 150, the
program will output news/admin/109 and news/admin/150.  One can then
remove the files listed in the output.

I don't know what the C news mkhistory does with duplicates.

I can email or post the program if there is a need.

				Jerry Aguirre