I have complicated feelings regarding #Usenet and #NNTP due to some server-centric aspects.
Granted due to message #gossip the death of any given instance isn't catastrophic and moving is largely unnoticeable, but it still puts some hurdles on usability in adverse conditions.
It still fulfills most of the #AsynchronousCommunication characteristics handily, but that's still a nagging thought, since /most/ instances demand fairly high-uptime to peer and don't allow such instability from peers.
@lispi314 not so. When I ran an NNTP-over-UUCP server I dialled in to my upstreams once every day (and later twice each day but both in off-peak time, so not at neat twelve hour intervals). My 'downstreams' dialled into me whenever they chose to but typically also in off-peak time, for cost reasons.
A continuous network connection is definitely not necessary.
Add comment