On Mon 27 Aug 2018 at 11:37:48 (-0400), Gene Heskett wrote: > On Monday 27 August 2018 11:11:37 David Wright wrote: > > > On Sat 25 Aug 2018 at 14:27:38 (-0500), John Hasler wrote: > > > David writes: > > > > Or are you talking about some type of "shared channel" of which I > > > > have no knowledge? > > > > > > Cable providers may have a great many customers on a single cable > > > with large (but limited) bandwidth. > > > > Oh, like me, you mean. When we wanted to get our cable strung from the > > pole with the least obstructed view of our house, the linesman first > > told us that all the terminations were taken, but on ringing the > > office, he found that one line was not subscribed to, so we were able > > to connect to that pole. When I walk down the back alleys, I can see > > other poles connected to the same main coax feed that links the poles. > > > > I'm still scratching my head why subscribing to NNTP newsgroups should > > lead to bandwidth problems rather than usage ones. I can hit my > > bandwidth limits in many other ways like downloading youtube videos, > > watching TV, etc, but the hard limit is my usage, where I would end up > > paying money for any excess. > > > That bandwidth limit is not on your side of the isp, its the bandwidth > from the main trunk lines to the isp. NNTP is a huge bandwidth hog > regardless of how much of it your isp accepts for spooling on local disk > to serve you.
I didn't know they were asking the ISP to *host* the newsgroups, just to allow NNTP stuff to pass from whoever is hosting it to the user, who pays for the usage they make of it. > > > Some rural providers may have limited backhaul bandwidth. They make > > > promises to customers based on optimistic estimates of peak usage. > Here at least, thats gradually getting better. I read that improvements are very patchy in the US. > > Now it appears that you're using "usage" where I would write > > "bandwidth". Am I in a minority of one here? Bandwidth is the rate of > > transfer of bits, whereas usage is the quantity of bits transferred > > irrespective of how fast they accumulated. > > Thats a pretty good view of the differences. But I get the impression that we have another baud/bitrate muddle. Cheers, David.
