David Lawrence <tale@uunet.uu.net> explains:
The problem is almost always interaction between NNTP and C News. NNTP POST asks C News's inews to not background itself but rather hang around and give its exit status so it knows whether the post was successful. (That wait will on some systems not return the exit status of the waited for job is a different sort of problem.) It ends up taking a long time because inews is calling relaynews, which often waits for another relaynews to free the lock on the news system so it can file the article. My preferred solution is to change inews to not call relaynews, but rather use newsspool. This loses some error-catching functionality, but is for the most part safe as inews will detect a lot of the errors on its own. The C News folks have sped up inews, too, so speed should look better to most folks as that update propagates around.