-- [ Picked text/plain from multipart/alternative ] Ahhh I worked it out hy some of you aren't some of you aren't getting the updates you should be getting. 1. tickrate is a command line option so set in your startup cmd line -tickrate 66 or -tickrate 100 2. You must ping boost your server by changing the Windows Multimedia timer. The most fool proof method is to start Windows Media Player with clock.avi loaded but not playing. 3. sv_maxupdaterate has to be higher than the tickrate for tickrates to be reflected correctly in the net_graph 3 as per Jesper S�rensen's explanation *AND* you must either change the sv_maxupdaterate value in your server.cfgfile and reboot the server or change it manually and then change the map. Most sv_ variables can be changed via rcon but then a map change must be done for the values to feed through. Moreover, please do an rcon sv_maxupdaterate with no value to get the server to report back to you exactly what it thinks it is using, rather than assume what you think it is using. Thats about as much as I have worked out thus far. Where to go from here? Thanks On 5/18/05, Jesper S�rensen <[EMAIL PROTECTED]> wrote: > > Sounds unlikely....At least I've never observed that on any client > computer, and I've seen a few in my time :) > > You known that the figures your a looking for, are on the right hand > side (xx.x/s) ? (The left hand side is packet size) > > net_maxfragments has nothing to do with it - The updates/s that are > counted by netgraph,are game updates/s not packet/s - So if a game > updates spans 100 packets, it still only counts as 1 update. > > I wouldnt mock about with net_maxfragments too much. Although small > packets usually are favorites to traverse internet routers faster than > big packets. > Most packets from the Source netcode are small as it is, and the added > overhead of fragmenting packets is hardly worth it. > You only really want to change it if you are on a segmet where you > cannot send 1500 bytes packets, without some device having to fragment > them, in that case it would be prudent to set it (GRE tunnel etc). > > In from left to right: > - Size of last update (received) > - Avg. bw used/s for last 30/s (not sure about the 30, but I seem to > recall something about 30 from the SDK) > - Avg game updates/s for last 30/s > > Out from left to right: > - Size of last update (sent) > - Avg. bw used/s for last 30/s > - Avg game updates/s for last 30/s > > /Jesper > > > SxRxRnR wrote: > > > Whisper, > > I know for a fact that the cl_updaterate and cl_cmdrate to not directly > > correlate to the same values in the netgraph. I can have both values on > > default of 20 and 30 on my client and still get 70-100 for the in and > out > > values in the netgraph. You can try changing net_maxfragments to see if > > that changes anything (def. Max fragment bytes per packet). Try setting > > that to different values to see if anything changes. Has anyone seen > > higher > > values? > > > > SxRxRnR > > > > > > > > _______________________________________________ > > To unsubscribe, edit your list preferences, or view the list archives, > > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlds > > > > . > > > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlds > --
_______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
