On Fri, May 02, 2003 at 09:35:42AM -0700, Tom Holub wrote: > On Fri, May 02, 2003 at 09:15:51AM -0700, Trent Piepho wrote: > > I'd really like to see the packets per second on a per player basis > > for both input and output, in terms of say one second peak. I've > > always wondered what the auto-firing borgs that are popular now do > > to network bandwidth. If the client sends 100 fire phaser requests > > per second, does the server respond with 100 "phaser not ready" > > messages? Does spamming the server with change course packets > > cause other players' ntservs to not get any cputime? netrek > > doesn't request any real-time scheduling, so there's no reason the > > OS won't decide to sacrifice latency for throughput when > > scheduling. > You only get one update per tick. Warning messages, for example, are sent immediately, often as packets on their own. I expect that 40 phaser requests would result in 39+ warning packets as Trent is theorizing. That said, 39+ packets isn't much unless you are connecting to netrek through a dialup modem... Trent: 'spamming the server' would only allow you to grab cpu time reserved for your ntserv process. Eventually, your process gets swapped out, and everybody else gets a turn. Could this cause noticable lag? Maybe, but probably not. Each ntserv process is likely limited to a time slice of less than 0.1 seconds if it isn't calling system calls, and less if it is doing read()/write() operations, especially if the write() blocks because the outgoing system buffer is full (something else I assume would eventually happen if the 'spamming' is resulting in numerous response packets). mark -- mark at mielke.cc/markm at ncf.ca/markm at nortelnetworks.com __________________________ . . _ ._ . . .__ . . ._. .__ . . . .__ | Neighbourhood Coder |\/| |_| |_| |/ |_ |\/| | |_ | |/ |_ | | | | | | \ | \ |__ . | | .|. |__ |__ | \ |__ | Ottawa, Ontario, Canada One ring to rule them all, one ring to find them, one ring to bring them all and in the darkness bind them... http://mark.mielke.cc/ _______________________________________________ vanilla-list mailing list vanilla-list at us.netrek.org https://mailman.real-time.com/mailman/listinfo/vanilla-list