view docs/testing.txt @ 267:a53d3f2f2f38 noffle

[svn] * docs/noffle.lsm: Update
author bears
date Mon, 05 Aug 2002 23:15:49 +0100
parents 8bb62a391b83
children
line wrap: on
line source

SETTING UP A TEST ENVIRONMENT

last update: $Id: testing.txt 323 2001-11-11 04:20:28Z mirkol $
written by: Mirko Liß, Nov 2001


This document describes how to set up a testing environment for the unstable
developer version of NOFFLE on a linux box.

You might not want to connect to the outside world with your newest hack.
It's much better to install a stable version of NOFFLE and let it feed
postings to the developer version.

The procedure step-by-step:

1. Install a stable version of noffle as usual. Or INN, if you have to
   blemish your soul by straying from the flock. Use this server to get
   articles from the common usenet and to act as an upstream server for
   your hacked version of noffle.

2. Install your hack with:
    ./configure --enable-debug=yes --with-spooldir=$ANYWHERE/spool/noffle \
        --with-configfile=$ANYWHERE/etc/noffle.conf \
        --with-docdir=$ANYWHERE/doc \
        --prefix=$ANYWHERE/usr/local --program-suffix=hack
    make
    make install
   Now the binary should be installed at:
    $ANYWHERE/local/bin/nofflehack
   Create a link to inews if you want to test that mode as well.

3. Edit $ANYWHERE/etc/noffle.conf:
    server localhost
    log-debug all
   If you changed the syslog ident or the facility (to LOG_USER)
   ( see: grep Log_init src/noffle.c ), you have to edit
   /etc/syslog.conf and eventually create new empty logfiles.


4. If you have to delete the database, just type:
    make install-data-local
   to get rid of the strange error messages.

5. If you want to use nofflehack in server mode, just add another port
   entry to /etc/services. Assigned Port numbers are listed at:
   http://www.iana.org/assignments/port-numbers
   Then add another line to /etc/inetd.conf (or whatever you use).
   Don't forget to tell your newsreader about the port number you chose.

If you want to tamper with the article database, there are various nice
gdbm tools available for Perl or Python. I often prefer shell-scripts
and used gdbm-primitive-1.15.tar.gz listed at http://freshmeat.net.
Be careful with file locking, though.