Re: [hlds] Re: sv_unlag and Ping
Firstly i would like to thank ned for all those commands and advice he has given. When i get i time to look at the server again i'll give them a test with my testing team (i've not been of the best health atm). Also to m0gley look mate i am and it works i happen to know that the average bandwith usage per player is in the regaion of 3 - 3.5 kilobytes p/s but hey if you guys say i can't run the server at all and i've got 6 players max running fime then i'm winning already. FYI this is only practice for a bigger paid for server my clan is goin to invest in and coz i'm the smart assed techie i got the job of server admin. and finally my system specs for you all to laugh at athlon 3000+ barton (fried it only runs at 1.3) 1GB ddr 333 corsair match pair ram windows xp pro 120GB 7200rpm hdd the system is in the middle of an upgrade and i do have a habit of playing in the server from the client on the same system (if that makes sense) Ned wrote: David, I 'm wrong on the FPS HLDS Servers use the CVAR *sys_ticrate* to control how many frames per second are rendered. *sys_ticrate 300 and open Medialayer.exe *this page explains why. http://steampowered.custhelp.com/cgi-bin/steampowered.cfg/php/enduser/std_adp.php?p_faqid=108 and sv_tickrate isn't used in 1.6, I host CS Source, got confused -tickrate 33 is in the command line of CSS BTW what cpu are you using, how much ram? Windows? This link will show you your true speed http://www.dslreports.com/ Ned Ned wrote: Hi David, run in the console to see whats going on netgraph X Shows the network graph. This will bring up a graph showing the performance of your internet connection. It will also tell you your Frames Per Second, the current amount of packet loss, and the amount of choke. Packet loss and Choke are one of the major causes of lag. Valid choices are 1 2 3 or 0 (off) click the link for more info http://home.covad.net/~k25125/SteamyThings/NetGraph_Steam.htm // max frame rate (60-1000) higher is better if your machine can handle it *sys_ticrate 300* // Maximum updates per second that the server will allow (default 60) // 60 for updaterate is LAN ONLY use 13 for internet hlds // 20 is default but will cut the maxplayers you can handle in 1/2 // for SRCDS Servers use 30 - you might be able to use 20 sv_maxupdaterate 20 try changing this down to 13 lowest // Minimum updates per second that the server will allow (default 10) // this is the minimum playable updaterate, leave this at 13 for srcds 10 hlds sv_minupdaterate 10 //Enables player lag compensation 0 - 1 sv_unlag 1 // Maximum lag compensation in seconds (min. 0.00 max. 1.00) sv_maxunlag 1 // Enable instanced baselines - Saves network overhead sv_instancebaselines 1 //Force server side preloading (default 0) sv_forcepreload 1 sv_minrate & sv_maxrate are for map and spray loading only they don't effect play You still won't see much improvement at 7 players, you mite be able to get 5 to work 128kbs / 5 = 25.5kbs per player this link will give you more info http://steampowered.custhelp.com/cgi-bin/steampowered.cfg/php/enduser/std_adp.php?p_faqid=108 Good Luck, Ned PS: you can always add bots so there is more players add as the last line in server.cfg exec bot.cfg put this in a text file, name it bot.cfg [word wrap off] List of bot commands: // bots bot_add //needed to add bots bot_quota 6 //# of bots bot_quota_mode fill //bot_quota_mode - If 'Fill', the server will adjust bots to keep # players in the game, where # is bot_quota (from above) Default: Normal bot_difficulty 2 //1 easy 2 normal 3 hard 4 expert bot_chatter normal//radio chatter bot_auto_follow 1 //follow humans bot_auto_vacate 0 //kick for human if too many players ie: over maxplayers bot_join_after_player 1 // if this is 0 then bots will play the map with out humans bot_defer_to_human 1 //bot_defer_to_human - Can bots complete objectives if there are humans on a team? Default: No 1 bot_allow_rogues 0//crazy bots bot_walk 0//walk only bot_join_team any // team t, ct, any bot_eco_limit 2000//limits their spending, leaving $ for humans on the same team bot_all_weapons 1 //the following commands are null if this is set to 1 bot_allow_grenades 1 bot_allow_pistols 1 bot_allow_sub_machine_guns 1 bot_allow_shotguns 1 bot_allow_rifles 1 bot_allow_snipers 1 bot_allow_machine_guns 1 David Williams wrote: well guys this is all very interesting to see you fight over a scientific fact but the whole reason we are even have this disscucion is coz i have a problem and i quote "i'm new to this and i've been reading this thread with interest as i have a 128 kilobit upload and a 1 megabit download on my connection and i want to run 8 players on my cs 1.6 server (currently 6) now my rates are set low to around 6000 but it get's really laggy when 7 ppl are playing. I also have 2 computers on the same lan as the server but it would seem that effects
Re: [hlds] Re: sv_unlag and Ping
David, I 'm wrong on the FPS HLDS Servers use the CVAR *sys_ticrate* to control how many frames per second are rendered. *sys_ticrate 300 and open Medialayer.exe *this page explains why. http://steampowered.custhelp.com/cgi-bin/steampowered.cfg/php/enduser/std_adp.php?p_faqid=108 and sv_tickrate isn't used in 1.6, I host CS Source, got confused -tickrate 33 is in the command line of CSS BTW what cpu are you using, how much ram? Windows? This link will show you your true speed http://www.dslreports.com/ Ned Ned wrote: Hi David, run in the console to see whats going on netgraph X Shows the network graph. This will bring up a graph showing the performance of your internet connection. It will also tell you your Frames Per Second, the current amount of packet loss, and the amount of choke. Packet loss and Choke are one of the major causes of lag. Valid choices are 1 2 3 or 0 (off) click the link for more info http://home.covad.net/~k25125/SteamyThings/NetGraph_Steam.htm // max frame rate (60-1000) higher is better if your machine can handle it *sys_ticrate 300* // Maximum updates per second that the server will allow (default 60) // 60 for updaterate is LAN ONLY use 13 for internet hlds // 20 is default but will cut the maxplayers you can handle in 1/2 // for SRCDS Servers use 30 - you might be able to use 20 sv_maxupdaterate 20 try changing this down to 13 lowest // Minimum updates per second that the server will allow (default 10) // this is the minimum playable updaterate, leave this at 13 for srcds 10 hlds sv_minupdaterate 10 //Enables player lag compensation 0 - 1 sv_unlag 1 // Maximum lag compensation in seconds (min. 0.00 max. 1.00) sv_maxunlag 1 // Enable instanced baselines - Saves network overhead sv_instancebaselines 1 //Force server side preloading (default 0) sv_forcepreload 1 sv_minrate & sv_maxrate are for map and spray loading only they don't effect play You still won't see much improvement at 7 players, you mite be able to get 5 to work 128kbs / 5 = 25.5kbs per player this link will give you more info http://steampowered.custhelp.com/cgi-bin/steampowered.cfg/php/enduser/std_adp.php?p_faqid=108 Good Luck, Ned PS: you can always add bots so there is more players add as the last line in server.cfg exec bot.cfg put this in a text file, name it bot.cfg [word wrap off] List of bot commands: // bots bot_add //needed to add bots bot_quota 6 //# of bots bot_quota_mode fill //bot_quota_mode - If 'Fill', the server will adjust bots to keep # players in the game, where # is bot_quota (from above) Default: Normal bot_difficulty 2 //1 easy 2 normal 3 hard 4 expert bot_chatter normal//radio chatter bot_auto_follow 1 //follow humans bot_auto_vacate 0 //kick for human if too many players ie: over maxplayers bot_join_after_player 1 // if this is 0 then bots will play the map with out humans bot_defer_to_human 1 //bot_defer_to_human - Can bots complete objectives if there are humans on a team? Default: No 1 bot_allow_rogues 0//crazy bots bot_walk 0//walk only bot_join_team any // team t, ct, any bot_eco_limit 2000//limits their spending, leaving $ for humans on the same team bot_all_weapons 1 //the following commands are null if this is set to 1 bot_allow_grenades 1 bot_allow_pistols 1 bot_allow_sub_machine_guns 1 bot_allow_shotguns 1 bot_allow_rifles 1 bot_allow_snipers 1 bot_allow_machine_guns 1 David Williams wrote: well guys this is all very interesting to see you fight over a scientific fact but the whole reason we are even have this disscucion is coz i have a problem and i quote "i'm new to this and i've been reading this thread with interest as i have a 128 kilobit upload and a 1 megabit download on my connection and i want to run 8 players on my cs 1.6 server (currently 6) now my rates are set low to around 6000 but it get's really laggy when 7 ppl are playing. I also have 2 computers on the same lan as the server but it would seem that effects the lag to (as if they were talkin to it from outside the network) now i would like to know what value's i should use to try and compensate for the realitivly low general connection speed. any thought's welcome but please try and be as descriptive as you can. i need to know exactly what to do " now can we get back to the point before i lose my temper again like i did yesterday (i get mad when i'm tryin to learn and the teacher isn't paying attention coz he/she is pleasuring him/herself when i'm nopt lookin) cheers guys ___ 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
Re: [hlds] Re: sv_unlag and Ping
David Williams wrote: "i'm new to this and i've been reading this thread with interest as i have a 128 kilobit upload and a 1 megabit download on my connection and i want to run 8 players on my cs 1.6 server You can't. My 1.6 server shows a peak of 4.5KB per player, per second outbound traffic. So: 4,500 bytes x 8 players = 36,000 bytes per second required. Your connection is 128Kb or 128,000 bits per second which is equal to 16,000 Bytes per second. Now if we use my 4,500 number for each player, and divide that into your 16,000 number, we get 3.55 players for your connection. This is why it makes no sense. I don't see how you can run more than 3 and a half players. now can we get back to the point before i lose my temper again like i did yesterday (i get mad when i'm tryin to learn and the teacher isn't paying attention coz he/she is pleasuring him/herself when i'm nopt lookin) Get mad all you want. You weren't being clear on your connection in your initial email and I still don't think you know. 128Kb upload is not enough to run 6 players on let alone 8. Now, you say you're running that and I believe you. But I don't see how that's possible with a 128Kb connection so you must have something higher. Connect to some 'fast' FTP server somewhere and start uploading a 2 or 3MB file. Use a client that shows what the speed of the transfer is and tell us the "sustained" rate that it shows you. I suppose you could you an online bandwidth test too but I'm always skeptable of those. Your question is dependent on giving us reliable information on your network connection. Don't bash us for trying to figure out what you're not telling us. -- - m0gely http://quake2.telestream.com/ Q2 | Q3A | Counter-strike ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
Hi David, run in the console to see whats going on netgraph X Shows the network graph. This will bring up a graph showing the performance of your internet connection. It will also tell you your Frames Per Second, the current amount of packet loss, and the amount of choke. Packet loss and Choke are one of the major causes of lag. Valid choices are 1 2 3 or 0 (off) click the link for more info http://home.covad.net/~k25125/SteamyThings/NetGraph_Steam.htm sv_tickrate 33 // max frame rate (60-1000) fps_max 60 // Maximum updates per second that the server will allow (default 60) // 60 for updaterate is LAN ONLY use 13 for internet hlds // 20 is default but will cut the maxplayers you can handle in 1/2 // for SRCDS Servers use 30 - you might be able to use 20 sv_maxupdaterate 20 try changing this down to 13 lowest // Minimum updates per second that the server will allow (default 10) // this is the minimum playable updaterate, leave this at 13 for srcds 10 hlds sv_minupdaterate 10 //Enables player lag compensation 0 - 1 sv_unlag 1 // Maximum lag compensation in seconds (min. 0.00 max. 1.00) sv_maxunlag 1 // Enable instanced baselines - Saves network overhead sv_instancebaselines 1 //Force server side preloading (default 0) sv_forcepreload 1 sv_minrate & sv_maxrate are for map and spray loading only they don't effect play You still won't see much improvement at 7 players, you mite be able to get 5 to work 128kbs / 5 = 25.5kbs per player this link will give you more info http://steampowered.custhelp.com/cgi-bin/steampowered.cfg/php/enduser/std_adp.php?p_faqid=108 Good Luck, Ned PS: you can always add bots so there is more players add as the last line in server.cfg exec bot.cfg put this in a text file, name it bot.cfg [word wrap off] List of bot commands: // bots bot_add //needed to add bots bot_quota 6 //# of bots bot_quota_mode fill //bot_quota_mode - If 'Fill', the server will adjust bots to keep # players in the game, where # is bot_quota (from above) Default: Normal bot_difficulty 2 //1 easy 2 normal 3 hard 4 expert bot_chatter normal//radio chatter bot_auto_follow 1 //follow humans bot_auto_vacate 0 //kick for human if too many players ie: over maxplayers bot_join_after_player 1 // if this is 0 then bots will play the map with out humans bot_defer_to_human 1 //bot_defer_to_human - Can bots complete objectives if there are humans on a team? Default: No 1 bot_allow_rogues 0//crazy bots bot_walk 0//walk only bot_join_team any // team t, ct, any bot_eco_limit 2000//limits their spending, leaving $ for humans on the same team bot_all_weapons 1 //the following commands are null if this is set to 1 bot_allow_grenades 1 bot_allow_pistols 1 bot_allow_sub_machine_guns 1 bot_allow_shotguns 1 bot_allow_rifles 1 bot_allow_snipers 1 bot_allow_machine_guns 1 David Williams wrote: well guys this is all very interesting to see you fight over a scientific fact but the whole reason we are even have this disscucion is coz i have a problem and i quote "i'm new to this and i've been reading this thread with interest as i have a 128 kilobit upload and a 1 megabit download on my connection and i want to run 8 players on my cs 1.6 server (currently 6) now my rates are set low to around 6000 but it get's really laggy when 7 ppl are playing. I also have 2 computers on the same lan as the server but it would seem that effects the lag to (as if they were talkin to it from outside the network) now i would like to know what value's i should use to try and compensate for the realitivly low general connection speed. any thought's welcome but please try and be as descriptive as you can. i need to know exactly what to do " now can we get back to the point before i lose my temper again like i did yesterday (i get mad when i'm tryin to learn and the teacher isn't paying attention coz he/she is pleasuring him/herself when i'm nopt lookin) cheers guys ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
Sprout, the unlag command is either on or off sv_unlag : 0 - 1: , "sv" : Enables player lag compensation use sv_maxunlag to set the amount of lag compensation, IMO though if your servers are working, don't change a lot of stuff at once or you could screw your self. sv_maxunlag 0 .1 - 1: , "sv" : Maximum lag compensation in seconds sv_instancebaselines 0 - 1:: Enable instanced baselines. Saves network overhead. netgraph X Show the network graph. This will bring up a graph showing the performance of your internet connection. It will also tell you your Frames Per Second, the current amount of packet loss, and the amount of choke. Packet loss and Choke are one of the major causes of lag. Valid choices are 1 2 3 or 0 (off). of course there a bunch of other rate tweaks just type in 'cvarlist' in the console. Ned sprout wrote: well my question was mostly cause I run 3 40 person servers they run fine but always looking to tweek them that little additional bit - Original Message - From: "David Williams" <[EMAIL PROTECTED]> To: Sent: Sunday, July 17, 2005 7:41 AM Subject: Re: [hlds] Re: sv_unlag and Ping and to answer sprout who seems to be on the ball it's whatever the default is mate. rate = 6000 cmdupdate_rate = 101 update_rate = 101 sprout wrote: its 1.0 so realisticly you should set it much higher if you have a nice connection like 100Mb? - Original Message - From: "Clayton Macleod" <[EMAIL PROTECTED]> To: Sent: Sunday, July 17, 2005 12:13 AM Subject: Re: [hlds] Re: sv_unlag and Ping 1.0 On 7/16/05, sprout <[EMAIL PROTECTED]> wrote: what is default sv_unlag? -- Clayton Macleod ___ 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 ___ 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
Re: [hlds] Re: sv_unlag and Ping
-- [ Picked text/plain from multipart/alternative ] sorry With sv_maxrate of 6000 you need approximately 48KB/s of upload bandwidth for 8 players On 7/19/05, Whisper <[EMAIL PROTECTED]> wrote: > > b is fo bit > B is for Bytes > 128Kilobit is 16KB/s Upload speeds > With sv_maxrate of 6000 you need approximately 36KB/s of upload bandwidth > for 6 players > With sv_maxrate of 6000 you need approximately 48KB/s of upload bandwidth > for 6 players > Your problem is you do not have enough bandwidth. > You need to divide 16KB/s by the number of players then multiple that > number by 1000 to get your bytes per second you can theorectically sustain > on your connection. > My advice is, not to bother, as 128Kb Upload speed is not enough to run a > server on. > Does that help? > On 7/19/05, David Williams <[EMAIL PROTECTED]> wrote: > > > > well guys this is all very interesting to see you fight over a > > scientific fact but the whole reason we are even have this disscucion is > > > > coz i have a problem and i quote > > > > "i'm new to this and i've been reading this thread with interest as i > > have a 128 kilobit upload and a 1 megabit download on my connection and > > i want to run 8 players on my cs 1.6 server (currently > > 6) now my rates are set low to around 6000 but it get's really laggy > > when 7 ppl are playing. I also have 2 computers on the same lan as the > > server but it would seem that effects the lag to (as if they were > > talkin to it from outside the network) now i would like to know what > > value's i should use to try and compensate for the realitivly low > > general connection speed. any thought's welcome but please try and be as > > descriptive as you can. i need to know exactly what to do " > > > > now can we get back to the point before i lose my temper again like i > > did yesterday (i get mad when i'm tryin to learn and the teacher isn't > > paying attention coz he/she is pleasuring him/herself when i'm nopt > > lookin) > > > > cheers guys > > > > James Tucker wrote: > > > > >Clayton is correct, yes I made a mistake previously, but as he said in > > >his e-mail, there is no debate. > > > > > >On 7/18/05, Hemminger Corey SrA 735 CES/CEUD > > ><[EMAIL PROTECTED]> wrote: > > > > > > > > >>It's not 1000MB or 1,000,000KB ect... Computers only work with powers > > of > > >>2 so you get, 1,2,4,8,16,32,64,128 ect.., it takes 8 bits to make a > > >>byte. Thus 4 is 2 to the power of 2 in binary 4 would be 0010. > > >> > > >> > > > > > >A computer might store it like that (it transmitted it like that). But > > >4 in 8-bit binary un-encoded should surely by 0100. Anyone for now > > >teaching endianess? Why not move onto a swift lesson on 2's complement > > >and IEE754 floats. > > > > > > > > > > > >>One > > >>Byte is all 8 binary digits grouped together. So 1MB is actually > > 1048KB > > >>which is 1,048,576 Bytes 2 to the power of 20. then you take that and > > >>multiply that by 8 = 8,388,608 bits, which is all the ones and zeros > > >>your modem has to transmit. KB and MB are just units of deviation like > > > > >>millimeter, centimeter, meter, kilometer. For simplicity they just > > round > > >>things down, especially because like Macleod said you get a little > > over > > >>head in the data. > > >> > > >>For the internet you can't have an IP digit greater than 255 because > > in > > >>an 8 bit octet it's . thus an IP of > > >>192.168.0.1<http://192.168.0.1/>is > > >>0011.00010101..1 each place in the binary > > represents > > >>the 1,2,4,8,32,64,124 so the first octet that's 192 says there is only > > > > >>1-124 and 1-64 added together gives 192. So now you have had a brief > > >>explanation on Binary and you understand a little bit of how those 1's > > >>and 0's work in computers. > > >> > > >> > > > > > >Thank you so much. > > > > > > > > > > > >>-Original Message- > > >>From: [EMAIL PROTECTED] > > >>[mailto: [EMAIL PROTECTED] On Behalf Of sprout > > >>Sent: Saturday, July 16, 2005 7:39 PM > > >>To: hlds@list.valvesoftware.com >
Re: [hlds] Re: sv_unlag and Ping
-- [ Picked text/plain from multipart/alternative ] b is fo bit B is for Bytes 128Kilobit is 16KB/s Upload speeds With sv_maxrate of 6000 you need approximately 36KB/s of upload bandwidth for 6 players With sv_maxrate of 6000 you need approximately 48KB/s of upload bandwidth for 6 players Your problem is you do not have enough bandwidth. You need to divide 16KB/s by the number of players then multiple that number by 1000 to get your bytes per second you can theorectically sustain on your connection. My advice is, not to bother, as 128Kb Upload speed is not enough to run a server on. Does that help? On 7/19/05, David Williams <[EMAIL PROTECTED]> wrote: > > well guys this is all very interesting to see you fight over a > scientific fact but the whole reason we are even have this disscucion is > coz i have a problem and i quote > > "i'm new to this and i've been reading this thread with interest as i > have a 128 kilobit upload and a 1 megabit download on my connection and > i want to run 8 players on my cs 1.6 server (currently > 6) now my rates are set low to around 6000 but it get's really laggy > when 7 ppl are playing. I also have 2 computers on the same lan as the > server but it would seem that effects the lag to (as if they were > talkin to it from outside the network) now i would like to know what > value's i should use to try and compensate for the realitivly low > general connection speed. any thought's welcome but please try and be as > descriptive as you can. i need to know exactly what to do " > > now can we get back to the point before i lose my temper again like i > did yesterday (i get mad when i'm tryin to learn and the teacher isn't > paying attention coz he/she is pleasuring him/herself when i'm nopt > lookin) > > cheers guys > > James Tucker wrote: > > >Clayton is correct, yes I made a mistake previously, but as he said in > >his e-mail, there is no debate. > > > >On 7/18/05, Hemminger Corey SrA 735 CES/CEUD > ><[EMAIL PROTECTED]> wrote: > > > > > >>It's not 1000MB or 1,000,000KB ect... Computers only work with powers of > >>2 so you get, 1,2,4,8,16,32,64,128 ect.., it takes 8 bits to make a > >>byte. Thus 4 is 2 to the power of 2 in binary 4 would be 0010. > >> > >> > > > >A computer might store it like that (it transmitted it like that). But > >4 in 8-bit binary un-encoded should surely by 0100. Anyone for now > >teaching endianess? Why not move onto a swift lesson on 2's complement > >and IEE754 floats. > > > > > > > >>One > >>Byte is all 8 binary digits grouped together. So 1MB is actually 1048KB > >>which is 1,048,576 Bytes 2 to the power of 20. then you take that and > >>multiply that by 8 = 8,388,608 bits, which is all the ones and zeros > >>your modem has to transmit. KB and MB are just units of deviation like > >>millimeter, centimeter, meter, kilometer. For simplicity they just round > >>things down, especially because like Macleod said you get a little over > >>head in the data. > >> > >>For the internet you can't have an IP digit greater than 255 because in > >>an 8 bit octet it's . thus an IP of > >>192.168.0.1<http://192.168.0.1>is > >>0011.00010101..1 each place in the binary represents > >>the 1,2,4,8,32,64,124 so the first octet that's 192 says there is only > >>1-124 and 1-64 added together gives 192. So now you have had a brief > >>explanation on Binary and you understand a little bit of how those 1's > >>and 0's work in computers. > >> > >> > > > >Thank you so much. > > > > > > > >>-Original Message- > >>From: [EMAIL PROTECTED] > >>[mailto:[EMAIL PROTECTED] On Behalf Of sprout > >>Sent: Saturday, July 16, 2005 7:39 PM > >>To: hlds@list.valvesoftware.com > >>Subject: Re: [hlds] Re: sv_unlag and Ping > >> > >>MB=mega byte Mb means mega bit ... thats where the confusion is its all > >>in the abreviation but clayton has it right as well as james just > >>diffeernt views but for the reasoning of the server I think its figured > >>in bits so clayton is altimatly right > >>- Original Message - > >>From: "Steve Dalberg" <[EMAIL PROTECTED]> > >>To: > >>Sent: Saturday, July 16, 2005 12:01 PM > >>Subject: Re: [hlds] Re: sv_unlag and Ping > >> > >> > >> > >> > >>>I'll second what Clayton says... 100b/
Re: [hlds] Re: sv_unlag and Ping
well guys this is all very interesting to see you fight over a scientific fact but the whole reason we are even have this disscucion is coz i have a problem and i quote "i'm new to this and i've been reading this thread with interest as i have a 128 kilobit upload and a 1 megabit download on my connection and i want to run 8 players on my cs 1.6 server (currently 6) now my rates are set low to around 6000 but it get's really laggy when 7 ppl are playing. I also have 2 computers on the same lan as the server but it would seem that effects the lag to (as if they were talkin to it from outside the network) now i would like to know what value's i should use to try and compensate for the realitivly low general connection speed. any thought's welcome but please try and be as descriptive as you can. i need to know exactly what to do " now can we get back to the point before i lose my temper again like i did yesterday (i get mad when i'm tryin to learn and the teacher isn't paying attention coz he/she is pleasuring him/herself when i'm nopt lookin) cheers guys James Tucker wrote: Clayton is correct, yes I made a mistake previously, but as he said in his e-mail, there is no debate. On 7/18/05, Hemminger Corey SrA 735 CES/CEUD <[EMAIL PROTECTED]> wrote: It's not 1000MB or 1,000,000KB ect... Computers only work with powers of 2 so you get, 1,2,4,8,16,32,64,128 ect.., it takes 8 bits to make a byte. Thus 4 is 2 to the power of 2 in binary 4 would be 0010. A computer might store it like that (it transmitted it like that). But 4 in 8-bit binary un-encoded should surely by 0100. Anyone for now teaching endianess? Why not move onto a swift lesson on 2's complement and IEE754 floats. One Byte is all 8 binary digits grouped together. So 1MB is actually 1048KB which is 1,048,576 Bytes 2 to the power of 20. then you take that and multiply that by 8 = 8,388,608 bits, which is all the ones and zeros your modem has to transmit. KB and MB are just units of deviation like millimeter, centimeter, meter, kilometer. For simplicity they just round things down, especially because like Macleod said you get a little over head in the data. For the internet you can't have an IP digit greater than 255 because in an 8 bit octet it's . thus an IP of 192.168.0.1 is 0011.00010101..1 each place in the binary represents the 1,2,4,8,32,64,124 so the first octet that's 192 says there is only 1-124 and 1-64 added together gives 192. So now you have had a brief explanation on Binary and you understand a little bit of how those 1's and 0's work in computers. Thank you so much. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of sprout Sent: Saturday, July 16, 2005 7:39 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping MB=mega byte Mb means mega bit ... thats where the confusion is its all in the abreviation but clayton has it right as well as james just diffeernt views but for the reasoning of the server I think its figured in bits so clayton is altimatly right - Original Message - From: "Steve Dalberg" <[EMAIL PROTECTED]> To: Sent: Saturday, July 16, 2005 12:01 PM Subject: Re: [hlds] Re: sv_unlag and Ping I'll second what Clayton says... 100b/s is 1Mbps Clayton Macleod wrote: sorry, but you're wrong. 1Mbps in terms of *network communication* is always 1,000,000 bits, just like 1Kbps is always 1,000 bits. On 7/16/05, James Tucker <[EMAIL PROTECTED]> wrote: Sorry, but I just want to verify, you do know those byte values are wrong don't you? 1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. Gb->Mb->Kb always factors of 1024 different. There are 8 bits in a byte. 1Mbps (bits per second, the standard measurement for most telcommunications speeds) 1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. 128k is actually 131072 bits per second 16k is 16384 bits per second. Rounded values are however good as they leave some space for oversubscription / link control / protocol overhead. Yeah, I couldn't recommend running a server on 16kbps up. -- Clayton Macleod ___ 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 __
Re: [hlds] Re: sv_unlag and Ping
Clayton is correct, yes I made a mistake previously, but as he said in his e-mail, there is no debate. On 7/18/05, Hemminger Corey SrA 735 CES/CEUD <[EMAIL PROTECTED]> wrote: > It's not 1000MB or 1,000,000KB ect... Computers only work with powers of > 2 so you get, 1,2,4,8,16,32,64,128 ect.., it takes 8 bits to make a > byte. Thus 4 is 2 to the power of 2 in binary 4 would be 0010. A computer might store it like that (it transmitted it like that). But 4 in 8-bit binary un-encoded should surely by 0100. Anyone for now teaching endianess? Why not move onto a swift lesson on 2's complement and IEE754 floats. > One > Byte is all 8 binary digits grouped together. So 1MB is actually 1048KB > which is 1,048,576 Bytes 2 to the power of 20. then you take that and > multiply that by 8 = 8,388,608 bits, which is all the ones and zeros > your modem has to transmit. KB and MB are just units of deviation like > millimeter, centimeter, meter, kilometer. For simplicity they just round > things down, especially because like Macleod said you get a little over > head in the data. > > For the internet you can't have an IP digit greater than 255 because in > an 8 bit octet it's . thus an IP of 192.168.0.1 is > 0011.00010101..1 each place in the binary represents > the 1,2,4,8,32,64,124 so the first octet that's 192 says there is only > 1-124 and 1-64 added together gives 192. So now you have had a brief > explanation on Binary and you understand a little bit of how those 1's > and 0's work in computers. Thank you so much. > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of sprout > Sent: Saturday, July 16, 2005 7:39 PM > To: hlds@list.valvesoftware.com > Subject: Re: [hlds] Re: sv_unlag and Ping > > MB=mega byte Mb means mega bit ... thats where the confusion is its all > in the abreviation but clayton has it right as well as james just > diffeernt views but for the reasoning of the server I think its figured > in bits so clayton is altimatly right > - Original Message - > From: "Steve Dalberg" <[EMAIL PROTECTED]> > To: > Sent: Saturday, July 16, 2005 12:01 PM > Subject: Re: [hlds] Re: sv_unlag and Ping > > > > I'll second what Clayton says... 100b/s is 1Mbps > > > > Clayton Macleod wrote: > > > >>sorry, but you're wrong. 1Mbps in terms of *network communication* is > >>always 1,000,000 bits, just like 1Kbps is always 1,000 bits. > >> > >>On 7/16/05, James Tucker <[EMAIL PROTECTED]> wrote: > >> > >> > >>>Sorry, but I just want to verify, you do know those byte values are > >>>wrong don't you? > >>> > >>>1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. > >>> > >>>Gb->Mb->Kb always factors of 1024 different. > >>>There are 8 bits in a byte. > >>> > >>>1Mbps (bits per second, the standard measurement for most > >>>telcommunications speeds) > >>> > >>>1Mbps is capable of sending 1024kbps, which is 1048576 bits per > second. > >>>128k is actually 131072 bits per second 16k is 16384 bits per second. > >>> > >>>Rounded values are however good as they leave some space for > >>>oversubscription / link control / protocol overhead. > >>> > >>>Yeah, I couldn't recommend running a server on 16kbps up. > >>> > >>> > >> > >> > >>-- > >>Clayton Macleod > >> > >>___ > >>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 > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
my word, you are a little mixed up. 1MB is 1024KB, not 1048KB, though you got the actual number of bytes correct. But I'm sorry to tell you that communications is indeed measured in base 10 numbers, not base 2. As I stated already, 1Mbps is 1,000,000 bits per second, and 1Kbps is 1,000 bits per second. Most definitely not 1,048,576 bits per second or 1024 bits per second or anything else like that. Yes, computers are binary devices. Yes, they only work with ones and zeroes. Yes, their math is all based off that fact. But communications speeds are all measured with literal thousands and millions, rather than base 2 mathematics, 1024 etc. A 56Kbps modem isn't capable of a maximum of 57344bps, it is capable of a maximum of 56000bps. (Telephone line voltage permitting...) This is well-established fact, not worth debating because there is no debate. On 7/18/05, Hemminger Corey SrA 735 CES/CEUD <[EMAIL PROTECTED]> wrote: > It's not 1000MB or 1,000,000KB ect... Computers only work with powers of > 2 so you get, 1,2,4,8,16,32,64,128 ect.., it takes 8 bits to make a > byte. Thus 4 is 2 to the power of 2 in binary 4 would be 0010. One > Byte is all 8 binary digits grouped together. So 1MB is actually 1048KB > which is 1,048,576 Bytes 2 to the power of 20. then you take that and > multiply that by 8 = 8,388,608 bits, which is all the ones and zeros > your modem has to transmit. KB and MB are just units of deviation like > millimeter, centimeter, meter, kilometer. For simplicity they just round > things down, especially because like Macleod said you get a little over > head in the data. > > For the internet you can't have an IP digit greater than 255 because in > an 8 bit octet it's . thus an IP of 192.168.0.1 is > 0011.00010101..1 each place in the binary represents > the 1,2,4,8,32,64,124 so the first octet that's 192 says there is only > 1-124 and 1-64 added together gives 192. So now you have had a brief > explanation on Binary and you understand a little bit of how those 1's > and 0's work in computers. -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Re: sv_unlag and Ping
It's not 1000MB or 1,000,000KB ect... Computers only work with powers of 2 so you get, 1,2,4,8,16,32,64,128 ect.., it takes 8 bits to make a byte. Thus 4 is 2 to the power of 2 in binary 4 would be 0010. One Byte is all 8 binary digits grouped together. So 1MB is actually 1048KB which is 1,048,576 Bytes 2 to the power of 20. then you take that and multiply that by 8 = 8,388,608 bits, which is all the ones and zeros your modem has to transmit. KB and MB are just units of deviation like millimeter, centimeter, meter, kilometer. For simplicity they just round things down, especially because like Macleod said you get a little over head in the data. For the internet you can't have an IP digit greater than 255 because in an 8 bit octet it's . thus an IP of 192.168.0.1 is 0011.00010101..1 each place in the binary represents the 1,2,4,8,32,64,124 so the first octet that's 192 says there is only 1-124 and 1-64 added together gives 192. So now you have had a brief explanation on Binary and you understand a little bit of how those 1's and 0's work in computers. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of sprout Sent: Saturday, July 16, 2005 7:39 PM To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping MB=mega byte Mb means mega bit ... thats where the confusion is its all in the abreviation but clayton has it right as well as james just diffeernt views but for the reasoning of the server I think its figured in bits so clayton is altimatly right - Original Message - From: "Steve Dalberg" <[EMAIL PROTECTED]> To: Sent: Saturday, July 16, 2005 12:01 PM Subject: Re: [hlds] Re: sv_unlag and Ping > I'll second what Clayton says... 100b/s is 1Mbps > > Clayton Macleod wrote: > >>sorry, but you're wrong. 1Mbps in terms of *network communication* is >>always 1,000,000 bits, just like 1Kbps is always 1,000 bits. >> >>On 7/16/05, James Tucker <[EMAIL PROTECTED]> wrote: >> >> >>>Sorry, but I just want to verify, you do know those byte values are >>>wrong don't you? >>> >>>1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. >>> >>>Gb->Mb->Kb always factors of 1024 different. >>>There are 8 bits in a byte. >>> >>>1Mbps (bits per second, the standard measurement for most >>>telcommunications speeds) >>> >>>1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. >>>128k is actually 131072 bits per second 16k is 16384 bits per second. >>> >>>Rounded values are however good as they leave some space for >>>oversubscription / link control / protocol overhead. >>> >>>Yeah, I couldn't recommend running a server on 16kbps up. >>> >>> >> >> >>-- >>Clayton Macleod >> >>___ >>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
Re: [hlds] Re: sv_unlag and Ping
On 7/16/05, Clayton Macleod <[EMAIL PROTECTED]> wrote: > sorry, but you're wrong. 1Mbps in terms of *network communication* is > always 1,000,000 bits, just like 1Kbps is always 1,000 bits. Yes, you're right. Clearly you're on the ball today ;) > > On 7/16/05, James Tucker <[EMAIL PROTECTED]> wrote: > > Sorry, but I just want to verify, you do know those byte values are > > wrong don't you? > > > > 1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. > > > > Gb->Mb->Kb always factors of 1024 different. > > There are 8 bits in a byte. > > > > 1Mbps (bits per second, the standard measurement for most > > telcommunications speeds) > > > > 1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. > > 128k is actually 131072 bits per second > > 16k is 16384 bits per second. > > > > Rounded values are however good as they leave some space for > > oversubscription / link control / protocol overhead. > > > > Yeah, I couldn't recommend running a server on 16kbps up. > > > -- > Clayton Macleod > > ___ > 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
Re: [hlds] Re: sv_unlag and Ping
well my question was mostly cause I run 3 40 person servers they run fine but always looking to tweek them that little additional bit - Original Message - From: "David Williams" <[EMAIL PROTECTED]> To: Sent: Sunday, July 17, 2005 7:41 AM Subject: Re: [hlds] Re: sv_unlag and Ping and to answer sprout who seems to be on the ball it's whatever the default is mate. rate = 6000 cmdupdate_rate = 101 update_rate = 101 sprout wrote: its 1.0 so realisticly you should set it much higher if you have a nice connection like 100Mb? - Original Message - From: "Clayton Macleod" <[EMAIL PROTECTED]> To: Sent: Sunday, July 17, 2005 12:13 AM Subject: Re: [hlds] Re: sv_unlag and Ping 1.0 On 7/16/05, sprout <[EMAIL PROTECTED]> wrote: what is default sv_unlag? -- Clayton Macleod ___ 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 ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
and to answer sprout who seems to be on the ball it's whatever the default is mate. rate = 6000 cmdupdate_rate = 101 update_rate = 101 sprout wrote: its 1.0 so realisticly you should set it much higher if you have a nice connection like 100Mb? - Original Message - From: "Clayton Macleod" <[EMAIL PROTECTED]> To: Sent: Sunday, July 17, 2005 12:13 AM Subject: Re: [hlds] Re: sv_unlag and Ping 1.0 On 7/16/05, sprout <[EMAIL PROTECTED]> wrote: what is default sv_unlag? -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
well i thought i was strange when the guy and tech support told me to divide the 1,024 value (the download speed 1 megabit and the upload 128 kilobits) i am currently runnin 6 players on any map very comfortably and i really don't apprieciate being treated like an idiot it's not that hard to figure out. so now lets get back to the point of my original question shall we. WHAT SHOULD I SET MY RATES TO (AND IGNORE THE FACT THE I SHOULDN'T DO IT YOPUR ONLY WASTING MY TIME) James Tucker wrote: Sorry, but I just want to verify, you do know those byte values are wrong don't you? 1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. Gb->Mb->Kb always factors of 1024 different. There are 8 bits in a byte. 1Mbps (bits per second, the standard measurement for most telcommunications speeds) 1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. 128k is actually 131072 bits per second 16k is 16384 bits per second. Rounded values are however good as they leave some space for oversubscription / link control / protocol overhead. Yeah, I couldn't recommend running a server on 16kbps up. On 7/15/05, m0gely <[EMAIL PROTECTED]> wrote: David Williams wrote: i'm new to this and i've been reading this thread with interest as i have a 16k upload and a 128k download on my connection (1 megabit fibre optic cable) 1Mb = 1,024,000 bytes 128k = 128,000 bytes 16k =16,000 bytes Something you're saying isn't making sense to me. So I'm just going to use your upload and download numbers. and i want to run 8 players on my cs 1.6 server I don't see how you could get more than 4 people on that connection without considerable lag. My server graphs indicate 4.5Kb per player outgoing as the ave for 1.6. So, if you're certain that 16k is in fact your outgoing limit, then this line just isn't suitable to run a server, unless you like very small maps for 4 player action. Outgoing line speed is where the bulk of the traffic is due to each client getting info from the server about all the other clients. -- - m0gely http://quake2.telestream.com/ Q2 | Q3A | Counter-strike ___ 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
Re: [hlds] Re: sv_unlag and Ping
its 1.0 so realisticly you should set it much higher if you have a nice connection like 100Mb? - Original Message - From: "Clayton Macleod" <[EMAIL PROTECTED]> To: Sent: Sunday, July 17, 2005 12:13 AM Subject: Re: [hlds] Re: sv_unlag and Ping 1.0 On 7/16/05, sprout <[EMAIL PROTECTED]> wrote: what is default sv_unlag? -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
1.0 On 7/16/05, sprout <[EMAIL PROTECTED]> wrote: > what is default sv_unlag? -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
what is default sv_unlag? - Original Message - From: "Mikee" <[EMAIL PROTECTED]> To: Sent: Saturday, July 16, 2005 9:25 PM Subject: Re: [hlds] Re: sv_unlag and Ping I am always amazed at how the simple concept of bits and bytes are confused by so many. It goes back to hardware engineers using the Metric (base 10) system in describing Hard Drive storage sizes, and bandwidth transfer speeds being expressed in "bits" of information. Then the software programmers use base 2, and only describe useable data expressed in terms of bytes (comprised of 8 bits--think of the parts of a digital clock showing the numeral 8 which has 7 pieces + 1 where they are all off). It is pretty confusing that the only real way to properly distinguish between bits and bytes is with a lower and upper case letter "b"/"B" respectively. They should have called a "Byte" something like a "Zoot" so there would never have been this confusion. - Original Message - From: "sprout" <[EMAIL PROTECTED]> To: Sent: Saturday, July 16, 2005 1:39 PM Subject: Re: [hlds] Re: sv_unlag and Ping MB=mega byte Mb means mega bit ... thats where the confusion is its all in the abreviation but clayton has it right as well as james just diffeernt views but for the reasoning of the server I think its figured in bits so clayton is altimatly right ___ 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
Re: [hlds] Re: sv_unlag and Ping
I am always amazed at how the simple concept of bits and bytes are confused by so many. It goes back to hardware engineers using the Metric (base 10) system in describing Hard Drive storage sizes, and bandwidth transfer speeds being expressed in "bits" of information. Then the software programmers use base 2, and only describe useable data expressed in terms of bytes (comprised of 8 bits--think of the parts of a digital clock showing the numeral 8 which has 7 pieces + 1 where they are all off). It is pretty confusing that the only real way to properly distinguish between bits and bytes is with a lower and upper case letter "b"/"B" respectively. They should have called a "Byte" something like a "Zoot" so there would never have been this confusion. - Original Message - From: "sprout" <[EMAIL PROTECTED]> To: Sent: Saturday, July 16, 2005 1:39 PM Subject: Re: [hlds] Re: sv_unlag and Ping MB=mega byte Mb means mega bit ... thats where the confusion is its all in the abreviation but clayton has it right as well as james just diffeernt views but for the reasoning of the server I think its figured in bits so clayton is altimatly right ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
sprout wrote: MB=mega byte Mb means mega bit bit's is what I meant, yes I know that a bad typographical error to make. -- - m0gely http://quake2.telestream.com/ Q2 | Q3A | Counter-strike ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
MB=mega byte Mb means mega bit ... thats where the confusion is its all in the abreviation but clayton has it right as well as james just diffeernt views but for the reasoning of the server I think its figured in bits so clayton is altimatly right - Original Message - From: "Steve Dalberg" <[EMAIL PROTECTED]> To: Sent: Saturday, July 16, 2005 12:01 PM Subject: Re: [hlds] Re: sv_unlag and Ping I'll second what Clayton says... 100b/s is 1Mbps Clayton Macleod wrote: sorry, but you're wrong. 1Mbps in terms of *network communication* is always 1,000,000 bits, just like 1Kbps is always 1,000 bits. On 7/16/05, James Tucker <[EMAIL PROTECTED]> wrote: Sorry, but I just want to verify, you do know those byte values are wrong don't you? 1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. Gb->Mb->Kb always factors of 1024 different. There are 8 bits in a byte. 1Mbps (bits per second, the standard measurement for most telcommunications speeds) 1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. 128k is actually 131072 bits per second 16k is 16384 bits per second. Rounded values are however good as they leave some space for oversubscription / link control / protocol overhead. Yeah, I couldn't recommend running a server on 16kbps up. -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
I'll second what Clayton says... 100b/s is 1Mbps Clayton Macleod wrote: sorry, but you're wrong. 1Mbps in terms of *network communication* is always 1,000,000 bits, just like 1Kbps is always 1,000 bits. On 7/16/05, James Tucker <[EMAIL PROTECTED]> wrote: Sorry, but I just want to verify, you do know those byte values are wrong don't you? 1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. Gb->Mb->Kb always factors of 1024 different. There are 8 bits in a byte. 1Mbps (bits per second, the standard measurement for most telcommunications speeds) 1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. 128k is actually 131072 bits per second 16k is 16384 bits per second. Rounded values are however good as they leave some space for oversubscription / link control / protocol overhead. Yeah, I couldn't recommend running a server on 16kbps up. -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
sorry, but you're wrong. 1Mbps in terms of *network communication* is always 1,000,000 bits, just like 1Kbps is always 1,000 bits. On 7/16/05, James Tucker <[EMAIL PROTECTED]> wrote: > Sorry, but I just want to verify, you do know those byte values are > wrong don't you? > > 1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. > > Gb->Mb->Kb always factors of 1024 different. > There are 8 bits in a byte. > > 1Mbps (bits per second, the standard measurement for most > telcommunications speeds) > > 1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. > 128k is actually 131072 bits per second > 16k is 16384 bits per second. > > Rounded values are however good as they leave some space for > oversubscription / link control / protocol overhead. > > Yeah, I couldn't recommend running a server on 16kbps up. -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
Sorry, but I just want to verify, you do know those byte values are wrong don't you? 1MB is 1024 KB which is 1048576 Bytes, which is 8388608 bits. Gb->Mb->Kb always factors of 1024 different. There are 8 bits in a byte. 1Mbps (bits per second, the standard measurement for most telcommunications speeds) 1Mbps is capable of sending 1024kbps, which is 1048576 bits per second. 128k is actually 131072 bits per second 16k is 16384 bits per second. Rounded values are however good as they leave some space for oversubscription / link control / protocol overhead. Yeah, I couldn't recommend running a server on 16kbps up. On 7/15/05, m0gely <[EMAIL PROTECTED]> wrote: > David Williams wrote: > > > i'm new to this and i've been reading this thread with interest as i > > have a 16k upload and a 128k download on my connection (1 megabit fibre > > optic cable) > > 1Mb = 1,024,000 bytes > 128k = 128,000 bytes > 16k =16,000 bytes > > Something you're saying isn't making sense to me. So I'm just going to > use your upload and download numbers. > > > and i want to run 8 players on my cs 1.6 server > > I don't see how you could get more than 4 people on that connection > without considerable lag. My server graphs indicate 4.5Kb per player > outgoing as the ave for 1.6. So, if you're certain that 16k is in fact > your outgoing limit, then this line just isn't suitable to run a server, > unless you like very small maps for 4 player action. Outgoing line > speed is where the bulk of the traffic is due to each client getting > info from the server about all the other clients. > > -- > - m0gely > http://quake2.telestream.com/ > Q2 | Q3A | Counter-strike > > ___ > 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
Re: [hlds] Re: sv_unlag and Ping
David Williams wrote: i'm new to this and i've been reading this thread with interest as i have a 16k upload and a 128k download on my connection (1 megabit fibre optic cable) 1Mb = 1,024,000 bytes 128k = 128,000 bytes 16k =16,000 bytes Something you're saying isn't making sense to me. So I'm just going to use your upload and download numbers. and i want to run 8 players on my cs 1.6 server I don't see how you could get more than 4 people on that connection without considerable lag. My server graphs indicate 4.5Kb per player outgoing as the ave for 1.6. So, if you're certain that 16k is in fact your outgoing limit, then this line just isn't suitable to run a server, unless you like very small maps for 4 player action. Outgoing line speed is where the bulk of the traffic is due to each client getting info from the server about all the other clients. -- - m0gely http://quake2.telestream.com/ Q2 | Q3A | Counter-strike ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
James Tucker wrote: IMHO cl_interpolate could be locked, but i dont really see the need, what does need to be done is to ensure that all players (ideally) are playing the same game tick at the same time. cl_interp should be locked at 0.1 or lower, any longer starts to get quite noticeable. I play on 0.05 regularly right now, but I am a broadband user. cl_smooth surely should default to 0, or in the very least not be FORCED to 1. I'd like to know: cl_lagcomp_errorcheck, does it work? it feels good, but that may just be a placebo. can't see anything on showhitboxes or showimpacts. On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: I agree completely, cl_interp and cl_interpolate should be locked or at least a cheat cvar. I can see no reason why you should be able to change them apart from for debugging purposes perhaps. Alfred can you shed any light on this? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of James Tucker Sent: 28 June 2005 12:23 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping I thought as clayton did for quite some time, and the reason is simply due to the use of the word "interpolation" and the fact that some documentation (alhtough I cant find it) suggested cl_interp was a 'maximum' value. The documentation at the link posted explains that the value of cl_interp is a delay value, by which all game world representation is re-wound prior to rendering. Unfortuantely, this means that the fight is for lower values of cl_interp (you will see your enemies before they see you). I don't know many other peoples direct opinions but I know that it's possible to die in less than 50ms. It does beg the question though, why cl_interp is not locked for internet gaming. I would imagine it is similar however to the state of cl_smooth and the choices of default updaterate and cmdrate values. So much for cheats really, when you can clearly gain advantage by optimising your cvars. On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: That's not how it works I'm afraid ;) It's all explained in this guide: http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html It explains the interpolation, prediction, and lag compensation. Regards, Ben -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: 27 June 2005 17:59 To: hlds@list.valvesoftware.com Subject: RE: [hlds] Re: sv_unlag and Ping I'm not totally sure about this, so anyone can feel free to correct me. I think if you turn cl_interpolate of, but set cl_interp 0.5, then the hitbox lags .5 seconds behinds the player. Some players might use this to hit players that appear to have already hidden behind a wall, or run through a crack in the doors... I've never tried it, as I try to get my models and hitboxes in the same place, but thats what I've kind of gathered. Again, not totally sure on this... Thanks for input though On Monday, June 27, 2005 2:01 am, Ben said: What are you worried about exactly? People using high interp values to cause annoyance to other players? Seems like a strange thing to worry about. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg Sent: 26 June 2005 23:24 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping Would turning sv_unlag down to the highest acceptable limit on our server keep people from using interpolate to have hitboxes lag behind models? We usually accept a 125ms-140ms ping, and start kicking at higher than that... (mainly as a way for the regs to get in). So should we set sv_unlag to 0.140? Thanks, Steve iceflatline wrote: well, obviously... Clayton Macleod wrote: well, obviously, sv_unlag is what turns on the compensation for clients' latency. It's what makes it so you don't have to lead your shots in front of the model you see, so you can actually aim at the model as if you had no latency at all. Turning it off just brings you back to the quake 1 days when there was nothing done to compensate for latency at all. The only situation where you might consider turning this off is on a LAN, but even then it makes no sense to turn it off, really. On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: i agree. i run hl and tfc servers (iceflatline.homeip.net) with that cvar enabled and it does seem to help clients with latency issues. -- Clayton Macleod ___ 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 __
Re: [hlds] Re: sv_unlag and Ping
IMHO cl_interpolate could be locked, but i dont really see the need, what does need to be done is to ensure that all players (ideally) are playing the same game tick at the same time. cl_interp should be locked at 0.1 or lower, any longer starts to get quite noticeable. I play on 0.05 regularly right now, but I am a broadband user. cl_smooth surely should default to 0, or in the very least not be FORCED to 1. I'd like to know: cl_lagcomp_errorcheck, does it work? it feels good, but that may just be a placebo. can't see anything on showhitboxes or showimpacts. On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: > I agree completely, cl_interp and cl_interpolate should be locked or at > least a cheat cvar. > > > I can see no reason why you should be able to change them apart from for > debugging purposes perhaps. > > Alfred can you shed any light on this? > > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of James Tucker > Sent: 28 June 2005 12:23 > To: hlds@list.valvesoftware.com > Subject: Re: [hlds] Re: sv_unlag and Ping > > I thought as clayton did for quite some time, and the reason is simply > due to the use of the word "interpolation" and the fact that some > documentation (alhtough I cant find it) suggested cl_interp was a > 'maximum' value. > > The documentation at the link posted explains that the value of > cl_interp is a delay value, by which all game world representation is > re-wound prior to rendering. Unfortuantely, this means that the fight > is for lower values of cl_interp (you will see your enemies before > they see you). I don't know many other peoples direct opinions but I > know that it's possible to die in less than 50ms. It does beg the > question though, why cl_interp is not locked for internet gaming. I > would imagine it is similar however to the state of cl_smooth and the > choices of default updaterate and cmdrate values. > > So much for cheats really, when you can clearly gain advantage by > optimising your cvars. > > On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: > > That's not how it works I'm afraid ;) It's all explained in this guide: > > http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html > > > > It explains the interpolation, prediction, and lag compensation. > > Regards, > > Ben > > > > -Original Message- > > From: [EMAIL PROTECTED] > > [mailto:[EMAIL PROTECTED] On Behalf Of > [EMAIL PROTECTED] > > Sent: 27 June 2005 17:59 > > To: hlds@list.valvesoftware.com > > Subject: RE: [hlds] Re: sv_unlag and Ping > > > > I'm not totally sure about this, so anyone can feel free to correct me. I > > think if you turn cl_interpolate of, but set cl_interp 0.5, then the > > hitbox lags .5 seconds behinds the player. Some players might use this to > > hit players that appear to have already hidden behind a wall, or run > > through a crack in the doors... I've never tried it, as I try to get my > > models and hitboxes in the same place, but thats what I've kind of > > gathered. > > > > Again, not totally sure on this... > > > > Thanks for input though > > On Monday, June 27, 2005 2:01 am, Ben said: > > > What are you worried about exactly? People using high interp values to > > > cause > > > annoyance to other players? Seems like a strange thing to worry about. > > > > > > -Original Message- > > > From: [EMAIL PROTECTED] > > > [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg > > > Sent: 26 June 2005 23:24 > > > To: hlds@list.valvesoftware.com > > > Subject: Re: [hlds] Re: sv_unlag and Ping > > > > > > Would turning sv_unlag down to the highest acceptable limit on our > > > server keep people from using interpolate to have hitboxes lag behind > > > models? We usually accept a 125ms-140ms ping, and start kicking at > > > higher than that... (mainly as a way for the regs to get in). So should > > > we set sv_unlag to 0.140? > > > > > > Thanks, > > > Steve > > > > > > > > > iceflatline wrote: > > > > > > > well, obviously... > > > > > > > > Clayton Macleod wrote: > > > > > > > >> well, obviously, sv_unlag is what turns on the compensation for > > > >> clients' latency. It's what makes it so you don't have to lead your > > > >> shots in front of the model you see, so you can actually aim at the > > > >> model as if you had no latency at all. Tu
RE: [hlds] Re: sv_unlag and Ping
Cl_interp is not machine/network specific - the default of 0.1 was chosen because it minimises the effects of lost packets. There is no reason why anyone should be allowed to set it any lower. I suspect however that if valve did remove the ability to change it that there would be an uproar from the community as many misinformed people believe that setting cl_interp to 0 "lines up the hitboxes with the models" or some other silly witchcraft. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of James Tucker Sent: 28 June 2005 12:23 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping I thought as clayton did for quite some time, and the reason is simply due to the use of the word "interpolation" and the fact that some documentation (alhtough I cant find it) suggested cl_interp was a 'maximum' value. The documentation at the link posted explains that the value of cl_interp is a delay value, by which all game world representation is re-wound prior to rendering. Unfortuantely, this means that the fight is for lower values of cl_interp (you will see your enemies before they see you). I don't know many other peoples direct opinions but I know that it's possible to die in less than 50ms. It does beg the question though, why cl_interp is not locked for internet gaming. I would imagine it is similar however to the state of cl_smooth and the choices of default updaterate and cmdrate values. So much for cheats really, when you can clearly gain advantage by optimising your cvars. On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: > That's not how it works I'm afraid ;) It's all explained in this guide: > http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html > > It explains the interpolation, prediction, and lag compensation. > Regards, > Ben > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] > Sent: 27 June 2005 17:59 > To: hlds@list.valvesoftware.com > Subject: RE: [hlds] Re: sv_unlag and Ping > > I'm not totally sure about this, so anyone can feel free to correct me. I > think if you turn cl_interpolate of, but set cl_interp 0.5, then the > hitbox lags .5 seconds behinds the player. Some players might use this to > hit players that appear to have already hidden behind a wall, or run > through a crack in the doors... I've never tried it, as I try to get my > models and hitboxes in the same place, but thats what I've kind of > gathered. > > Again, not totally sure on this... > > Thanks for input though > On Monday, June 27, 2005 2:01 am, Ben said: > > What are you worried about exactly? People using high interp values to > > cause > > annoyance to other players? Seems like a strange thing to worry about. > > > > -Original Message- > > From: [EMAIL PROTECTED] > > [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg > > Sent: 26 June 2005 23:24 > > To: hlds@list.valvesoftware.com > > Subject: Re: [hlds] Re: sv_unlag and Ping > > > > Would turning sv_unlag down to the highest acceptable limit on our > > server keep people from using interpolate to have hitboxes lag behind > > models? We usually accept a 125ms-140ms ping, and start kicking at > > higher than that... (mainly as a way for the regs to get in). So should > > we set sv_unlag to 0.140? > > > > Thanks, > > Steve > > > > > > iceflatline wrote: > > > > > well, obviously... > > > > > > Clayton Macleod wrote: > > > > > >> well, obviously, sv_unlag is what turns on the compensation for > > >> clients' latency. It's what makes it so you don't have to lead your > > >> shots in front of the model you see, so you can actually aim at the > > >> model as if you had no latency at all. Turning it off just brings you > > >> back to the quake 1 days when there was nothing done to compensate for > > >> latency at all. The only situation where you might consider turning > > >> this off is on a LAN, but even then it makes no sense to turn it off, > > >> really. > > >> > > >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > > >> > > >> > > >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > > >>> cvar enabled and it does seem to help clients with latency issues. > > >>> > > >>> > > >> > > >> > > >> -- > > >> Clayton Macleod > > >> > > >>
RE: [hlds] Re: sv_unlag and Ping
I agree completely, cl_interp and cl_interpolate should be locked or at least a cheat cvar. I can see no reason why you should be able to change them apart from for debugging purposes perhaps. Alfred can you shed any light on this? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of James Tucker Sent: 28 June 2005 12:23 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping I thought as clayton did for quite some time, and the reason is simply due to the use of the word "interpolation" and the fact that some documentation (alhtough I cant find it) suggested cl_interp was a 'maximum' value. The documentation at the link posted explains that the value of cl_interp is a delay value, by which all game world representation is re-wound prior to rendering. Unfortuantely, this means that the fight is for lower values of cl_interp (you will see your enemies before they see you). I don't know many other peoples direct opinions but I know that it's possible to die in less than 50ms. It does beg the question though, why cl_interp is not locked for internet gaming. I would imagine it is similar however to the state of cl_smooth and the choices of default updaterate and cmdrate values. So much for cheats really, when you can clearly gain advantage by optimising your cvars. On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: > That's not how it works I'm afraid ;) It's all explained in this guide: > http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html > > It explains the interpolation, prediction, and lag compensation. > Regards, > Ben > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] > Sent: 27 June 2005 17:59 > To: hlds@list.valvesoftware.com > Subject: RE: [hlds] Re: sv_unlag and Ping > > I'm not totally sure about this, so anyone can feel free to correct me. I > think if you turn cl_interpolate of, but set cl_interp 0.5, then the > hitbox lags .5 seconds behinds the player. Some players might use this to > hit players that appear to have already hidden behind a wall, or run > through a crack in the doors... I've never tried it, as I try to get my > models and hitboxes in the same place, but thats what I've kind of > gathered. > > Again, not totally sure on this... > > Thanks for input though > On Monday, June 27, 2005 2:01 am, Ben said: > > What are you worried about exactly? People using high interp values to > > cause > > annoyance to other players? Seems like a strange thing to worry about. > > > > -Original Message- > > From: [EMAIL PROTECTED] > > [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg > > Sent: 26 June 2005 23:24 > > To: hlds@list.valvesoftware.com > > Subject: Re: [hlds] Re: sv_unlag and Ping > > > > Would turning sv_unlag down to the highest acceptable limit on our > > server keep people from using interpolate to have hitboxes lag behind > > models? We usually accept a 125ms-140ms ping, and start kicking at > > higher than that... (mainly as a way for the regs to get in). So should > > we set sv_unlag to 0.140? > > > > Thanks, > > Steve > > > > > > iceflatline wrote: > > > > > well, obviously... > > > > > > Clayton Macleod wrote: > > > > > >> well, obviously, sv_unlag is what turns on the compensation for > > >> clients' latency. It's what makes it so you don't have to lead your > > >> shots in front of the model you see, so you can actually aim at the > > >> model as if you had no latency at all. Turning it off just brings you > > >> back to the quake 1 days when there was nothing done to compensate for > > >> latency at all. The only situation where you might consider turning > > >> this off is on a LAN, but even then it makes no sense to turn it off, > > >> really. > > >> > > >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > > >> > > >> > > >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > > >>> cvar enabled and it does seem to help clients with latency issues. > > >>> > > >>> > > >> > > >> > > >> -- > > >> Clayton Macleod > > >> > > >> ___ > > >> To unsubscribe, edit your list preferences, or view the list > > >> archives, please visit: > > >> http://list.valvesoftware.com/mailman/listinfo/hlds > > >
Re: [hlds] Re: sv_unlag and Ping
I thought as clayton did for quite some time, and the reason is simply due to the use of the word "interpolation" and the fact that some documentation (alhtough I cant find it) suggested cl_interp was a 'maximum' value. The documentation at the link posted explains that the value of cl_interp is a delay value, by which all game world representation is re-wound prior to rendering. Unfortuantely, this means that the fight is for lower values of cl_interp (you will see your enemies before they see you). I don't know many other peoples direct opinions but I know that it's possible to die in less than 50ms. It does beg the question though, why cl_interp is not locked for internet gaming. I would imagine it is similar however to the state of cl_smooth and the choices of default updaterate and cmdrate values. So much for cheats really, when you can clearly gain advantage by optimising your cvars. On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: > That's not how it works I'm afraid ;) It's all explained in this guide: > http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html > > It explains the interpolation, prediction, and lag compensation. > Regards, > Ben > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] > Sent: 27 June 2005 17:59 > To: hlds@list.valvesoftware.com > Subject: RE: [hlds] Re: sv_unlag and Ping > > I'm not totally sure about this, so anyone can feel free to correct me. I > think if you turn cl_interpolate of, but set cl_interp 0.5, then the > hitbox lags .5 seconds behinds the player. Some players might use this to > hit players that appear to have already hidden behind a wall, or run > through a crack in the doors... I've never tried it, as I try to get my > models and hitboxes in the same place, but thats what I've kind of > gathered. > > Again, not totally sure on this... > > Thanks for input though > On Monday, June 27, 2005 2:01 am, Ben said: > > What are you worried about exactly? People using high interp values to > > cause > > annoyance to other players? Seems like a strange thing to worry about. > > > > -Original Message- > > From: [EMAIL PROTECTED] > > [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg > > Sent: 26 June 2005 23:24 > > To: hlds@list.valvesoftware.com > > Subject: Re: [hlds] Re: sv_unlag and Ping > > > > Would turning sv_unlag down to the highest acceptable limit on our > > server keep people from using interpolate to have hitboxes lag behind > > models? We usually accept a 125ms-140ms ping, and start kicking at > > higher than that... (mainly as a way for the regs to get in). So should > > we set sv_unlag to 0.140? > > > > Thanks, > > Steve > > > > > > iceflatline wrote: > > > > > well, obviously... > > > > > > Clayton Macleod wrote: > > > > > >> well, obviously, sv_unlag is what turns on the compensation for > > >> clients' latency. It's what makes it so you don't have to lead your > > >> shots in front of the model you see, so you can actually aim at the > > >> model as if you had no latency at all. Turning it off just brings you > > >> back to the quake 1 days when there was nothing done to compensate for > > >> latency at all. The only situation where you might consider turning > > >> this off is on a LAN, but even then it makes no sense to turn it off, > > >> really. > > >> > > >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > > >> > > >> > > >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > > >>> cvar enabled and it does seem to help clients with latency issues. > > >>> > > >>> > > >> > > >> > > >> -- > > >> Clayton Macleod > > >> > > >> ___ > > >> 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 prefere
RE: [hlds] Re: sv_unlag and Ping
Doh! There's an article written by yahn about it too that is floating about in various places on the net - http://www.gamesurge.com/pc/interviews/netcode.shtml - interview http://www.gdconf.com/archives/2001/bernier.doc - article -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Clayton Macleod Sent: 28 June 2005 11:54 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping that link's no good anymore :( On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: > That's not how it works I'm afraid ;) It's all explained in this guide: > http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html > > It explains the interpolation, prediction, and lag compensation. > Regards, > Ben -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
that link's no good anymore :( On 6/28/05, Ben <[EMAIL PROTECTED]> wrote: > That's not how it works I'm afraid ;) It's all explained in this guide: > http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html > > It explains the interpolation, prediction, and lag compensation. > Regards, > Ben -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Re: sv_unlag and Ping
That's not how it works I'm afraid ;) It's all explained in this guide: http://www.valve-erc.com/srcsdk/general/multiplayer_networking.html It explains the interpolation, prediction, and lag compensation. Regards, Ben -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: 27 June 2005 17:59 To: hlds@list.valvesoftware.com Subject: RE: [hlds] Re: sv_unlag and Ping I'm not totally sure about this, so anyone can feel free to correct me. I think if you turn cl_interpolate of, but set cl_interp 0.5, then the hitbox lags .5 seconds behinds the player. Some players might use this to hit players that appear to have already hidden behind a wall, or run through a crack in the doors... I've never tried it, as I try to get my models and hitboxes in the same place, but thats what I've kind of gathered. Again, not totally sure on this... Thanks for input though On Monday, June 27, 2005 2:01 am, Ben said: > What are you worried about exactly? People using high interp values to > cause > annoyance to other players? Seems like a strange thing to worry about. > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg > Sent: 26 June 2005 23:24 > To: hlds@list.valvesoftware.com > Subject: Re: [hlds] Re: sv_unlag and Ping > > Would turning sv_unlag down to the highest acceptable limit on our > server keep people from using interpolate to have hitboxes lag behind > models? We usually accept a 125ms-140ms ping, and start kicking at > higher than that... (mainly as a way for the regs to get in). So should > we set sv_unlag to 0.140? > > Thanks, > Steve > > > iceflatline wrote: > > > well, obviously... > > > > Clayton Macleod wrote: > > > >> well, obviously, sv_unlag is what turns on the compensation for > >> clients' latency. It's what makes it so you don't have to lead your > >> shots in front of the model you see, so you can actually aim at the > >> model as if you had no latency at all. Turning it off just brings you > >> back to the quake 1 days when there was nothing done to compensate for > >> latency at all. The only situation where you might consider turning > >> this off is on a LAN, but even then it makes no sense to turn it off, > >> really. > >> > >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > >> > >> > >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > >>> cvar enabled and it does seem to help clients with latency issues. > >>> > >>> > >> > >> > >> -- > >> Clayton Macleod > >> > >> ___ > >> 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 > > > ___ > 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
Re: [hlds] Re: sv_unlag and Ping
as I understand it the interp value is just a maximum, and it doesn't interpolate anywhere past the data it is receiving. So if you are receiving data every 50ms, 20 updates a second, but getting 40fps, 25ms, it's not going to be interpolating anything beyond 25ms anyways since you are getting real data for every second frame that gets rendered. Is that not how it works in practice? I thought interpolation only applied from through to the , and had nothing to do with your ping/latency in that regard. Just my theory, could be wrong. But I think the lag compensation is kind of seperate from packet-to-packet interpolation. On 6/27/05, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > I'm not totally sure about this, so anyone can feel free to correct me. I > think if you turn cl_interpolate of, but set cl_interp 0.5, then the > hitbox lags .5 seconds behinds the player. Some players might use this to > hit players that appear to have already hidden behind a wall, or run > through a crack in the doors... I've never tried it, as I try to get my > models and hitboxes in the same place, but thats what I've kind of > gathered. > > Again, not totally sure on this... > > Thanks for input though -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Re: sv_unlag and Ping
I'm not totally sure about this, so anyone can feel free to correct me. I think if you turn cl_interpolate of, but set cl_interp 0.5, then the hitbox lags .5 seconds behinds the player. Some players might use this to hit players that appear to have already hidden behind a wall, or run through a crack in the doors... I've never tried it, as I try to get my models and hitboxes in the same place, but thats what I've kind of gathered. Again, not totally sure on this... Thanks for input though On Monday, June 27, 2005 2:01 am, Ben said: > What are you worried about exactly? People using high interp values to > cause > annoyance to other players? Seems like a strange thing to worry about. > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg > Sent: 26 June 2005 23:24 > To: hlds@list.valvesoftware.com > Subject: Re: [hlds] Re: sv_unlag and Ping > > Would turning sv_unlag down to the highest acceptable limit on our > server keep people from using interpolate to have hitboxes lag behind > models? We usually accept a 125ms-140ms ping, and start kicking at > higher than that... (mainly as a way for the regs to get in). So should > we set sv_unlag to 0.140? > > Thanks, > Steve > > > iceflatline wrote: > > > well, obviously... > > > > Clayton Macleod wrote: > > > >> well, obviously, sv_unlag is what turns on the compensation for > >> clients' latency. It's what makes it so you don't have to lead your > >> shots in front of the model you see, so you can actually aim at the > >> model as if you had no latency at all. Turning it off just brings you > >> back to the quake 1 days when there was nothing done to compensate for > >> latency at all. The only situation where you might consider turning > >> this off is on a LAN, but even then it makes no sense to turn it off, > >> really. > >> > >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > >> > >> > >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > >>> cvar enabled and it does seem to help clients with latency issues. > >>> > >>> > >> > >> > >> -- > >> Clayton Macleod > >> > >> ___ > >> 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 > > > ___ > 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
RE: [hlds] Re: sv_unlag and Ping
What are you worried about exactly? People using high interp values to cause annoyance to other players? Seems like a strange thing to worry about. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg Sent: 26 June 2005 23:24 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping Would turning sv_unlag down to the highest acceptable limit on our server keep people from using interpolate to have hitboxes lag behind models? We usually accept a 125ms-140ms ping, and start kicking at higher than that... (mainly as a way for the regs to get in). So should we set sv_unlag to 0.140? Thanks, Steve iceflatline wrote: > well, obviously... > > Clayton Macleod wrote: > >> well, obviously, sv_unlag is what turns on the compensation for >> clients' latency. It's what makes it so you don't have to lead your >> shots in front of the model you see, so you can actually aim at the >> model as if you had no latency at all. Turning it off just brings you >> back to the quake 1 days when there was nothing done to compensate for >> latency at all. The only situation where you might consider turning >> this off is on a LAN, but even then it makes no sense to turn it off, >> really. >> >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: >> >> >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that >>> cvar enabled and it does seem to help clients with latency issues. >>> >>> >> >> >> -- >> Clayton Macleod >> >> ___ >> 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 ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
RE: [hlds] Re: sv_unlag and Ping
No, because that doesn't include the default interpolation. You need to do max ping + max interp value. So with default and 140 ping max = 0.14 + 0.1 = 0.24 -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Steve Dalberg Sent: 26 June 2005 23:24 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping Would turning sv_unlag down to the highest acceptable limit on our server keep people from using interpolate to have hitboxes lag behind models? We usually accept a 125ms-140ms ping, and start kicking at higher than that... (mainly as a way for the regs to get in). So should we set sv_unlag to 0.140? Thanks, Steve iceflatline wrote: > well, obviously... > > Clayton Macleod wrote: > >> well, obviously, sv_unlag is what turns on the compensation for >> clients' latency. It's what makes it so you don't have to lead your >> shots in front of the model you see, so you can actually aim at the >> model as if you had no latency at all. Turning it off just brings you >> back to the quake 1 days when there was nothing done to compensate for >> latency at all. The only situation where you might consider turning >> this off is on a LAN, but even then it makes no sense to turn it off, >> really. >> >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: >> >> >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that >>> cvar enabled and it does seem to help clients with latency issues. >>> >>> >> >> >> -- >> Clayton Macleod >> >> ___ >> 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 ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
I think that's the point of being able to adjust that value, yeah. So if you have some no-mind from Australia coming on your server with a 650ms ping he can't cause aggravation from someone being around the corner and down the hall and dying there cuz this laggy dude shot him, from his perspective, before the guy even goes around the corner. Setting it to just above your average pings should be just fine. sv_maxunlag, though, no? On 6/26/05, Steve Dalberg <[EMAIL PROTECTED]> wrote: > Would turning sv_unlag down to the highest acceptable limit on our > server keep people from using interpolate to have hitboxes lag behind > models? We usually accept a 125ms-140ms ping, and start kicking at > higher than that... (mainly as a way for the regs to get in). So should > we set sv_unlag to 0.140? > > Thanks, > Steve > > > iceflatline wrote: > > > well, obviously... > > > > Clayton Macleod wrote: > > > >> well, obviously, sv_unlag is what turns on the compensation for > >> clients' latency. It's what makes it so you don't have to lead your > >> shots in front of the model you see, so you can actually aim at the > >> model as if you had no latency at all. Turning it off just brings you > >> back to the quake 1 days when there was nothing done to compensate for > >> latency at all. The only situation where you might consider turning > >> this off is on a LAN, but even then it makes no sense to turn it off, > >> really. > >> > >> On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > >> > >> > >>> i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > >>> cvar enabled and it does seem to help clients with latency issues. > >>> > >>> > >> > >> > >> -- > >> Clayton Macleod > >> > >> ___ > >> 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 > -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
Would turning sv_unlag down to the highest acceptable limit on our server keep people from using interpolate to have hitboxes lag behind models? We usually accept a 125ms-140ms ping, and start kicking at higher than that... (mainly as a way for the regs to get in). So should we set sv_unlag to 0.140? Thanks, Steve iceflatline wrote: well, obviously... Clayton Macleod wrote: well, obviously, sv_unlag is what turns on the compensation for clients' latency. It's what makes it so you don't have to lead your shots in front of the model you see, so you can actually aim at the model as if you had no latency at all. Turning it off just brings you back to the quake 1 days when there was nothing done to compensate for latency at all. The only situation where you might consider turning this off is on a LAN, but even then it makes no sense to turn it off, really. On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: i agree. i run hl and tfc servers (iceflatline.homeip.net) with that cvar enabled and it does seem to help clients with latency issues. -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
well, obviously... Clayton Macleod wrote: well, obviously, sv_unlag is what turns on the compensation for clients' latency. It's what makes it so you don't have to lead your shots in front of the model you see, so you can actually aim at the model as if you had no latency at all. Turning it off just brings you back to the quake 1 days when there was nothing done to compensate for latency at all. The only situation where you might consider turning this off is on a LAN, but even then it makes no sense to turn it off, really. On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: i agree. i run hl and tfc servers (iceflatline.homeip.net) with that cvar enabled and it does seem to help clients with latency issues. -- Clayton Macleod ___ 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
RE: [hlds] Re: sv_unlag and Ping
And it's on by default so you don't really need it in the config either. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Clayton Macleod Sent: 25 June 2005 21:23 To: hlds@list.valvesoftware.com Subject: Re: [hlds] Re: sv_unlag and Ping well, obviously, sv_unlag is what turns on the compensation for clients' latency. It's what makes it so you don't have to lead your shots in front of the model you see, so you can actually aim at the model as if you had no latency at all. Turning it off just brings you back to the quake 1 days when there was nothing done to compensate for latency at all. The only situation where you might consider turning this off is on a LAN, but even then it makes no sense to turn it off, really. On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > cvar enabled and it does seem to help clients with latency issues. -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
well, obviously, sv_unlag is what turns on the compensation for clients' latency. It's what makes it so you don't have to lead your shots in front of the model you see, so you can actually aim at the model as if you had no latency at all. Turning it off just brings you back to the quake 1 days when there was nothing done to compensate for latency at all. The only situation where you might consider turning this off is on a LAN, but even then it makes no sense to turn it off, really. On 6/25/05, iceflatline <[EMAIL PROTECTED]> wrote: > i agree. i run hl and tfc servers (iceflatline.homeip.net) with that > cvar enabled and it does seem to help clients with latency issues. -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
Re: [hlds] Re: sv_unlag and Ping
i agree. i run hl and tfc servers (iceflatline.homeip.net) with that cvar enabled and it does seem to help clients with latency issues. Clayton Macleod wrote: playing without sv_unlag makes no sense at all, especially not over the internet. On 6/25/05, leo bounds <[EMAIL PROTECTED]> wrote: I saw someone talking about sv_unlag and ping performance. Is it a good idea to have sv_unlag 1 in all our server.cfg files as a general rule to help people with bad ping rates and if so would that be for all HL games, CS, DOD, CSS etc ? Thanks for info -- Clayton Macleod ___ 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
Re: [hlds] Re: sv_unlag and Ping
playing without sv_unlag makes no sense at all, especially not over the internet. On 6/25/05, leo bounds <[EMAIL PROTECTED]> wrote: > I saw someone talking about sv_unlag and ping > performance. Is it a good idea to have sv_unlag 1 in > all our server.cfg files as a general rule to help > people with bad ping rates and if so would that be for > all HL games, CS, DOD, CSS etc ? > > Thanks for info -- Clayton Macleod ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds