Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 11/03/84 (WLS Mods); site astrovax.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!allegra!princeton!astrovax!wls
From: wls@astrovax.UUCP (William L. Sebok)
Newsgroups: net.news.group
Subject: Re: Topology, connectivity, and garbage
Message-ID: <679@astrovax.UUCP>
Date: Sat, 2-Nov-85 01:07:07 EST
Article-I.D.: astrovax.679
Posted: Sat Nov  2 01:07:07 1985
Date-Received: Sun, 3-Nov-85 06:30:59 EST
References: <787@adobe.UUCP>
Reply-To: wls@astrovax.UUCP (William L. Sebok)
Organization: Princeton Univ. Astrophysics
Lines: 19

In article <787@adobe.UUCP> greid@adobe.UUCP (Glenn Reid) writes:
>Another thought:  have an "option" to news that allows a site to turn
>off a newsgroup with the following behavior:  Any article posted to it
>that hits that site will be returned to the sender informing him that
>the news got no farther than that particular site.  That should discourage
>people from posting the value of pi.  Or any article > n bytes.  Or
>whatever.

This cure would be worse than the disease.  I can envision a flood of such
returned articles overwhelming the mail links.  If the article is mailed back
along the arrival path then a site that does this article returning has just
doubled the cost of the article to all of the intermediate sites.  If instead
the article is mailed along an optimized path: the average mail link tends to
be more expensive than the average news link (which is more often local), and
the article returning site has just flooded some expensive mail links with news
articles.
-- 
Bill Sebok			Princeton University, Astrophysics
{allegra,akgua,cbosgd,decvax,ihnp4,noao,philabs,princeton,vax135}!astrovax!wls