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 ucbvax.ARPA
Path: utzoo!linus!decvax!decwrl!ucbvax!fair
From: fair@ucbvax.ARPA (Erik E. Fair)
Newsgroups: net.news
Subject: Re: Why digests are bad
Message-ID: <10382@ucbvax.ARPA>
Date: Sat, 14-Sep-85 09:22:06 EDT
Article-I.D.: ucbvax.10382
Posted: Sat Sep 14 09:22:06 1985
Date-Received: Sun, 15-Sep-85 05:17:43 EDT
References: <10220@ucbvax.ARPA> <764@vortex.UUCP> <1471@umcp-cs.UUCP>
Organization: University of California at Berkeley
Lines: 29

I get the distinct impression that everyone missed the point
(everyone, that is, except Larry Wall and Charlie Mangoe).

With the information contained in the header of a news article,
your wonderful whizzy user-interface can construct a digest of all
the articles posted on September 10th, 1985 if you want.

It is wrong for the user interfaces to have undigestifying code in
them, because they should not have to deal with two kinds of article
structure. And because there is more useful information in the header
of single netnews articles than there is in the header of a generic
digest article (no message-id, or references, ...), there are more
things you can do with single netnews articles.

THAT is why digests are bad.

	1) They force us to have crufty code to deal with them in
		the user-interfaces (which from a design point of
		view is wrong)
	
	2) They break existing and potential functionality otherwise
		provided by a netnews article header, by not having
		the requisite information.

Larry Wall is also correct in stating that sorting of netnews articles
is the province of rnews, and therefore refusing to provide such
capability in rn (among other things, that would slow it down).

	Erik E. Fair	ucbvax!fair	fair@ucbarpa.BERKELEY.EDU