Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 beta 3/9/83; site cae780.UUCP
Path: utzoo!linus!philabs!seismo!hao!kpno!amd70!cae780!chuqui
From: chuqui@cae780.UUCP
Newsgroups: net.news.adm
Subject: Re: This bothers me a bit, so...
Message-ID: <258@cae780.UUCP>
Date: Wed, 28-Dec-83 18:07:59 EST
Article-I.D.: cae780.258
Posted: Wed Dec 28 18:07:59 1983
Date-Received: Fri, 30-Dec-83 00:39:23 EST
References: <4197@amd70.UUCP>, <743@ulysses.UUCP>
Organization: the Warlocks cave
Lines: 27


    Having been involved in trying to track down a net problem
    (articles were vanishing down a black hole), I can categorically
    state that there is a need for net.test.  We (unc!bch, unc!bts, and
    myself) were trying to determine why net.* articles from UNC
    weren't getting very far.  Using 'to.*' doesn't help.  Nor, it
    turns out, did using net.test, since many sites weren't propogating
    it.  In the end, we resorted to net.misc for a test message -- not
    out of ignorance, not out of malice, but simply because that was
    the only way we had to duplicate the problem.

This is the very reason why I have been sending out my messages. There are
some cases where a problem in the net needs to be tracked down, and that is
why net.test exists. Unfortunately, too many sites use it without thinking,
which causes other sites to stop forwarding it, which means that it isn't
useful for net testing. In MOST cases, passing messages back and forth to
your neighbors is all you need, and all you should do.

As far as the cost of my mail replies, I am sending a single message to a
person with some information he may not have known or remembered. If that
message keeps that person from shipping out a net.test message in the
future, the overall amount of traffic is reduced. The mail I have been
getting back has shown me that there are a large number of people who
hadn't known about alternatives to net.test (and thanked me for cluing them
in), so at this point I think its worth it...