I have a minor, sporadic problem with nntpcache-1.0, used on a dial-on-demand PPP link to the news-provider. Sometimes, the news-server at the news-provider will be down, or the PPP connection will fail for some reason. Nntpcache records the failure to get the active file, but records the `update' as finished. Feb 12 06:05:03 auda nntpcached-update[1038]: checking server news.connix.com for 'active' Feb 12 06:05:03 auda nntpcached-update[1038]: No children: list active on news.connix.com error: 503 Timeout after 7200 seconds, closing connection.^M Feb 12 06:05:03 auda nntpcached-update[1038]: update daemon finished Feb 12 06:05:03 auda ./nntpcached[1037]: connect from ron@auda (192.101.32.4) Feb 12 06:05:05 auda ./nntpcached[1037]: posting <lambing_855745503@mlfarm.com>^M Feb 12 06:06:20 auda ./nntpcached[1037]: Connection timed out: could not connect to news.connix.com as DEFAULT Feb 12 06:06:20 auda ./nntpcached[1037]: Connection timed out: couldn't post article to news.connix.com (server down) If another connection is made to nntpcache before the update interval for active (4h) has expired, `active' is not retrieved. For some newsreaders, this presents a problem. Is there a configuration or code change that can be made so a _failed_ effort to update the active file will result in a new effort at the next connection, _before_ the update interval has expired? Thanks, -- Ronald Florence Maple Lawn Farm, Stonington, CT ron@mlfarm.com http://www.connix.com/~mlfarm
