Thats why it was a counterpoint - these are centralized systems to the point that I would imagine any increase in traffic requires a ROI. :-)
-David Waite James MacMillan wrote: > Could you please clarify this for me? > By my understanding, this ping/pong technique only generates extra > traffic between the client and their > direct server (virtually a single hop). Since jabber is intended to be > a highly distributed network where > many users administer their own servers, ~14Mb of extra traffic spread > across the 100s of 1000s of servers > that would be serving those 5 million online users doesn't seem as > relevant. > > David Waite wrote: > >> >> >> David Waite wrote: >> >>> >>> As a counterpoint, however - a single-byte application-level >>> ping/pong is still going to take a minimum of 82 bytes of network >>> traffic. It is likely that for five million online users, the ~7MB/s >>> traffic a once-a-minute ping/pong would generate might not >>> considered worth it. >> >> >> >> Whoops, messed up my own numbers - it would be a minimum of 162, or >> ~14 MB/s >> >> _______________________________________________ >> jdev mailing list >> [EMAIL PROTECTED] >> http://mailman.jabber.org/listinfo/jdev > > > > > > _______________________________________________ > jdev mailing list > [EMAIL PROTECTED] > http://mailman.jabber.org/listinfo/jdev _______________________________________________ jdev mailing list [EMAIL PROTECTED] http://mailman.jabber.org/listinfo/jdev
