Looks like suburbia.net is on-hold... Can the mailing list be fixed to send out headers with nntpcache-users@nntpcache.org? Evan -----Original Message----- From: <MAILER-DAEMON@conductor.synapse.net> To: <evanc@synapse.net> Date: Sunday, August 16, 1998 5:15 PM Subject: failure notice >Hi. This is the qmail-send program at conductor.synapse.net. >I'm afraid I wasn't able to deliver your message to the following addresses. >This is a permanent error; I've given up. Sorry it didn't work out. > ><nntpcache-users@suburbia.net>: >Sorry, I couldn't find any host named suburbia.net. (#5.1.2) > ><nntpcache-users@suburbia.net>: >Sorry, I couldn't find any host named suburbia.net. (#5.1.2) > >--- Below this line is a copy of the message. > >Return-Path: <evanc@synapse.net> >Received: (qmail 28877 invoked from network); 16 Aug 1998 21:14:58 -0000 >Received: from localhost (HELO cello) (127.0.0.1) > by localhost with SMTP; 16 Aug 1998 21:14:58 -0000 >Message-ID: <014701bdc95a$c8a46ae0$c9252fce@synapse.net> >From: "Evan Champion" <evanc@synapse.net> >To: "Tom" <tom@sdf.com>, > <nntpcache-users@suburbia.net> >Cc: <nntpcache-users@suburbia.net> >Subject: Re: NNTPC: 2 servers for 1 group >Date: Sun, 16 Aug 1998 17:13:57 -0400 >MIME-Version: 1.0 >Content-Type: text/plain; > charset="iso-8859-1" >Content-Transfer-Encoding: 7bit >X-Priority: 3 >X-MSMail-Priority: Normal >X-Mailer: Microsoft Outlook Express 5.00.0518.4 >X-MimeOLE: Produced By Microsoft MimeOLE V5.00.0518.4 > >> Except basically all article handling will have to be done by >>message-IDs. This will probably slows things down, specially stuff like >>XOVER merging. > >Remember, however, that nntpcache's history file is much smaller than INN's >would be to host a spool covering the same number of days, so you shouldn't >have the same penalties for article retrieval by Message-ID. > >Perhaps a more significant change would be storage by Message-ID. I'd >personally really like to see nntpcache do cyclic article storage because it >does not seem to be very good at all at expiring... > >I don't know how feasible it would be to do this sort of thing in realtime. >Lots of Message-ID lookups may not be too bad as long as the hashtable can >be held in core. There are a lot of ways that the history file could be >reduced in size as well; for example, with a cyclic spool article paths may >be reduced to a fixed length file:offset:length, and on one of the history >file formats for INN they don't actually store the Message-ID but instead on >a subset of the Message-ID's MD5 hash (IIRC it was 5 of the 16 bytes), on >the basis that it is statistically extremely unlikely to have collisions, >and if there was a collision, well, too bad :-) Squid is using this >technique as well; instead of doing lookups on a URL, it looks up on the MD5 >hash of the URL. > >Evan > > >
