#how do i test network commands for performance and latency
1 messages · Page 1 of 1 (latest)
yes but these command dont get referenced or documented enough
thats why i ask if anyone has info about this and if theres a way i can test these
well
there is no reason to turn off compression or prediction im p sure
they are documented in mastercomfig and which options should be used
how come spy has cl_pred_optimize 1 previous versions of mastercomfig have net_compresspackets 0
like ive read the github but i just dont see any thing regarding whether why these are bad or why they should be left at default
compresspackets is just generally better for your network and everything i mean it just costs extra cpu power for less network load which helps with stuff like dropped packets
well its not wrong to have it on either so its not a bad setting
*well its not wrong to
have it on either so its
not a bad setting*
and from my experience makes mouse feel responsive
not sure if thats placebo but if i can test it ill do that
network settings dont effect the mouse at all it is 100% predicted (even when prediction is off)
im mean mouse responsiveness prob not the thing that changed but i feel that it is more responsive
i just dont know what it is or how i can quantify
You won't notice any of them. The only reason you'd need to choose something other than mastercomfig's defaults is if your internet connection is having issues.
regarding cl_pred_optimize 1 for spy: #1237735772166295553 message
no point in disabling packet compression, the min size already covers not compressing usercmds, net_queued_packet_thread 0 doesn't do anything to usercmds because they aren't split, net_splitpacket_maxrate doesn't do anything because usercmds dont split and if you did have them split, splitpacket rate must be extremely low in order to avoid them getting out of order
increasing rate is fine but thats a module in mastercomfig called bandwidth
cl_pred_optimize 1 will usually fuck with your rocket jumping
okak