wmartin@brl-vgr.ARPA (Will Martin ) (03/12/84)
If my system is down for some time, or if the machine to which this host connects in order to receive netnews is either down or somehow unreachable, is the netnews traffic which was circulated through that part of USENET during that time forever lost to me? It seems that I have often seen followup notes to an original submission which has never appeared. (I don't mean the normal leapfrogging where you see the answer before the question, but situations where the answers or replies show up for many days, but the original item never appears in our local files.) [Of course, I can never know when it is a followup that doesn't appear, unless someone else refers to it!] What traffic is normally made available to a system when it calls up to get new netnews? Just traffic which has recently arrived during the past "n" hours at the called site? Everything which has arrived since the calling machine last called? Everything on-line at the called site, no matter how old? (If this latter, it would be inefficient, but nobody would ever miss anything unless they were off-net for more than two weeks at a stretch, so I am pretty sure this isn't the case!) Do the normal netnews extraction/distribution processes only look for articles somehow characterized as "recent" or "new"? If so, do the system administrators have the capability to perform some special version of the netnews routines (after more than a day or so of separation from the net) which will cause the software to look for missed articles? Does this require special coordination between the staff at both the called and calling sites, or can it be done by action at the calling site alone? What about news items submitted from the isolated site? Are they all queued and sent out when a connection is finally made, or could they be "lost" (that is, not sent out to the net) if the originating system does not make a net connection for some period of time after the article is submitted? If so, what is this critical time period? I appreciate any explanations or descriptions of the processes involved which would make this entire operation clear to me. Regards, Will Martin