Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site sdcsvax.UUCP
Path: utzoo!watmath!clyde!cbosgd!ulysses!bellcore!decvax!ittatc!dcdwest!sdcsvax!brian
From: brian@sdcsvax.UUCP (Brian Kantor)
Newsgroups: net.news
Subject: Re: Statistics, polls: lan-news surveys
Message-ID: <1614@sdcsvax.UUCP>
Date: Fri, 28-Mar-86 13:41:22 EST
Article-I.D.: sdcsvax.1614
Posted: Fri Mar 28 13:41:22 1986
Date-Received: Sun, 30-Mar-86 10:41:41 EST
References: <2015@hao.UUCP> <3389@sun.uucp> <1365@mit-eddie.MIT.EDU> <1960@saber.UUCP> <12727@ucbvax.BERKELEY.EDU>
Reply-To: brian@sdcsvax.UUCP (Brian Kantor)
Distribution: net
Organization: UCSD wombat breeding society
Lines: 29

As one of the authors of the NNTP protocol (RFC977), I can speak with
some authority on this point: the survey process WILL NOT INCLUDE users
of the network news reading programs, if it is run on the system where
they read news.  The reason is that there is NO /usr/lib/news/active
file on that system - in fact, there doesn't have to be a /usr/lib/news
directory at all.

True, the users each have .newsrc files that do tally their
subscription activity, but the current news survey also requires access
to the 'active file, so it will not give valid results, if, indeed, it
will run at all.

Worse, many of the systems that converted over to use NNTP and rrn may
have left their old /usr/lib/news directories in place, with an old and
completely out-of-date active files.  I've not seen what actually
happens in this case.

Now that I've said all that, realize that the number of hosts running
nntp and rrn is really small.  I doubt that it would have much of any
effect on the statistics gathered from any significant number of hosts
using the conventional news system.

	Brian Kantor	UCSD Office of Academic Computing
			Academic Network Operations Group  
			UCSD B-028, La Jolla, CA 92093 (619) 452-6865

	decvax\ 	brian@sdcsvax.ucsd.edu
	ihnp4  >---  sdcsvax  --- brian
	ucbvax/		Kantor@Nosc