On Wed, 26 Jun 1996, Luke Bowker wrote: > If it works correctly connecting directly to the server then I suspect the > problem is in nntpcached. It may not be a problem at all though, since > nntpcache is designed to cache the active file for the period specified > in the nntpcache.servers file. When it receives a connect from a client, > it checks to see if the timeout interval has expired, and if it has, then > it hands out the OLD data, then forks off into the background to refetch > and collate the active file, this is to prevent any client from having to > wait for a time consuming active file rebuild. You may wish to check to > see if this is what's happening (syslog will tell you when it does rebuilds). > Does tin use the active file to check for unread messages in this mode? I > know that there's a few newsreader clients (eg. netscape), that check for > new news by simply sending group commands for each subscribed group. Group > commands aren't cached (to maintain synchronisation with the server(s)), > so tin might be finding new messages that the active file doesn't (yet) > show this way. This might be the issue, it seems to be so that with '-n' tin only issues group-commands and so it finds new news correctly. And when doing refresh group command is also used. Without '-n' tin retrieves the whole active file and doesn't find all new messages. My new problem is that 'nntpcached' sometimes terminates silently from the background (has done so 3 times now within 30 hours of operation) and I have to restart it. I'll investigate... --- Hannu Laurila - kube@japo.fi * Kauppakatu 10, FIN-62900 ALAJÄRVI Alajärven Puhelinosuuskunta * Tel +358 66 557 2209 - Fax +358 66 557 2788
