pmove_msec 5/10 and etpro ?
Moderators: Forum moderators, developers
pmove_msec 5/10 and etpro ?
I have bee fidling with a server 2 get et super smooth...
Im getting som funny results... 3.2.6 beta version...
Is pmove_msec fixed at 8ms i etpro? or can it be set and work serverside...
Trying 2 get it running at 5 ms or 10 ms intervals, where the logic would be that the server frames are locked at sv_fps 20 so the sv_frame is 50 ms... thereby u "should" get 10 or 5 "position" updates pr server frame.... Realy trying to find a best "matchup".. or "fillrate" of the updates so they add even up to match the server frame time of 50 ms...
Or am i looking at something thats bad juju 4 w.r.t server setup.. there is probaly a good reason its set at 8 ms.... default..
Any help/input apreciated...
Im getting som funny results... 3.2.6 beta version...
Is pmove_msec fixed at 8ms i etpro? or can it be set and work serverside...
Trying 2 get it running at 5 ms or 10 ms intervals, where the logic would be that the server frames are locked at sv_fps 20 so the sv_frame is 50 ms... thereby u "should" get 10 or 5 "position" updates pr server frame.... Realy trying to find a best "matchup".. or "fillrate" of the updates so they add even up to match the server frame time of 50 ms...
Or am i looking at something thats bad juju 4 w.r.t server setup.. there is probaly a good reason its set at 8 ms.... default..
Any help/input apreciated...
Super .. so pmove_msec is not hardwired in etpro...
If pmove_fixed 0 is set server and client side and pmove_msec is set 5,8 or 10 serverside what are the implications? would there be any changes in the server calculations? Is it still used at a base number in position calculations?
pmove_fixed 1 and pmove_mesc 10 set on the server and with a forcecvar pmove_fixed 1 also set, i have noticed that the sniper rifle in scoped mode is on target the recoil "bug" is gone... but im not sure it holds true concering all com_maxfps..i tried it at 100...
The buggy issues with pmove_fixed 1, cant they be solved by setting the pmove_msec 5 or 10 serverside?
One more thing.. at com_maxfps 100, pmove_fixed 1 and pmove_msec 10 set client side.. on a server at pmove_fixed 1 and pmove_msec 10 set server side... Its realy smooth and responsive.. even if if have my fps drop below 100...?
So what reasoning is behind choice of pmove_msec 8 as default value? Range is 3-33 right?...
pmove_msec can it be set fraction? ie. 4.5 or 8.44 ?
If pmove_fixed 0 is set server and client side and pmove_msec is set 5,8 or 10 serverside what are the implications? would there be any changes in the server calculations? Is it still used at a base number in position calculations?
pmove_fixed 1 and pmove_mesc 10 set on the server and with a forcecvar pmove_fixed 1 also set, i have noticed that the sniper rifle in scoped mode is on target the recoil "bug" is gone... but im not sure it holds true concering all com_maxfps..i tried it at 100...
The buggy issues with pmove_fixed 1, cant they be solved by setting the pmove_msec 5 or 10 serverside?
One more thing.. at com_maxfps 100, pmove_fixed 1 and pmove_msec 10 set client side.. on a server at pmove_fixed 1 and pmove_msec 10 set server side... Its realy smooth and responsive.. even if if have my fps drop below 100...?
So what reasoning is behind choice of pmove_msec 8 as default value? Range is 3-33 right?...
pmove_msec can it be set fraction? ie. 4.5 or 8.44 ?
I think it's 8 - 33.LilleBror wrote:Range is 3-33 right?...
8 was chosen since 1000 / 8 = 125 but I forget what was up with the 125 fps.
pmove_msec is an integer so no fractions.
Our servers now run on 64 bit steroids. Point your ET to:
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem
deej: any input on pmove_msec 5 or 10... i got a strange felling that since the server frame time is 50 ms... u "fill" the server frame better.. without "leftovers" and "exports" in the transition between serverframes..
It might even solve some prediction errors..between 2 recieved snaps?
my logic is 8ms * X = 50 ? no match in real/int..
50 cant use limmit is 33 (full)
25 * 2 = 50 ok! (1/2)
12.5 cant use fraction (1/4)
10 * 5 = 50 ok! (1/5)
8.333 cant use fraction (1/6)
6.25 cant use fraction (1/8)
5 * 10 = 50 ok! (1/10)
2 cant use limmit is 3?
if the server calulates like this: (8 ms) as 8+8+8+8+8+8+8 = 48...so in my logic 2 ms left... so what happens...
does the engine take a extra 8 check pr. 4 server frames ? or what?
...
5+8+8+8+8+8+5,? or 8+8+8+8+8+10 ?.. 9+8+8+8+8+9 ..mixed?..
...the agian it might be real trivial if pmove_msec has no effect if pmove_fixed is set 0 serverside...
is the pmove_msec a "base" even if pmove_fixed is set zero serverside...?
is it just me.. but i find it strange that the pmove_msec 8 (default)doesnt fit.. the standart sv_fps 20 timespace of 50 msec..
default 8 ms.. say use 5*8 = 40 ms.. *25 = 1000... so that should work best at sv_fps 25....?
but Etpro..is hardcoded at sv_fps 20... and stuff should break then?
..so.. maybe one could try with sv_fps 25? and se i pmove_fixed 1 set serverside still gives the sniperbug... at com_maxfps 125 set client side?
It might even solve some prediction errors..between 2 recieved snaps?
my logic is 8ms * X = 50 ? no match in real/int..
50 cant use limmit is 33 (full)
25 * 2 = 50 ok! (1/2)
12.5 cant use fraction (1/4)
10 * 5 = 50 ok! (1/5)
8.333 cant use fraction (1/6)
6.25 cant use fraction (1/8)
5 * 10 = 50 ok! (1/10)
2 cant use limmit is 3?
if the server calulates like this: (8 ms) as 8+8+8+8+8+8+8 = 48...so in my logic 2 ms left... so what happens...
does the engine take a extra 8 check pr. 4 server frames ? or what?
...
5+8+8+8+8+8+5,? or 8+8+8+8+8+10 ?.. 9+8+8+8+8+9 ..mixed?..
...the agian it might be real trivial if pmove_msec has no effect if pmove_fixed is set 0 serverside...
is the pmove_msec a "base" even if pmove_fixed is set zero serverside...?
is it just me.. but i find it strange that the pmove_msec 8 (default)doesnt fit.. the standart sv_fps 20 timespace of 50 msec..
default 8 ms.. say use 5*8 = 40 ms.. *25 = 1000... so that should work best at sv_fps 25....?
but Etpro..is hardcoded at sv_fps 20... and stuff should break then?
..so.. maybe one could try with sv_fps 25? and se i pmove_fixed 1 set serverside still gives the sniperbug... at com_maxfps 125 set client side?
sv_fps > 20 is not a good idea since it breaks stuff. There have been alot of posts by ReyalP, Rain & Bani on this. On pmove_sec 5: impossible, 8 is the lower limit AFAIK. pmove_msec 10 I have no idea but there must be a good reason why they default it to 8.
Our servers now run on 64 bit steroids. Point your ET to:
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem
yup.. sv_fps 25 "wierd" with the flammer...
--but--
gives the same "smooth" feel as pmove_msec 10 at pmove_fixed 1 and at com_maxfps 100 at sv_fps 20
with
pmove_msec 8 pmove_fixed 1 com_maxfps 125 at sv_fps 25
then agian...if its pr default broken... not an option..on a server...
"good old quake.. would run sv_fps 25 ? no problems...right?"
If its "only" wierd flammer graphics, covert ops uni steal time.. and timer pause issues... one could setup a test server..
But if the antilag, antiwarp or anything concering hit detection get messed up.. there is zero piont 2 it...
..i know that the unlagged code becomes less "effective" at high sv_fps..
Still the pmove_msec...hmm... u say 8 is min.. i got it set 3 and 5.. but "if its in effect on server-client"..thats something else? no? i have no idea if i can trust the console printout on that perticular issu..
But if 8 ms is the "must have" choice value and default... the maby one should try sv_fps 25? but im realy not positive .. most likely 99.9999% a complet waist of time...
I'd realy like some feedback on pmove_msec 10 set serverside...
--but--
gives the same "smooth" feel as pmove_msec 10 at pmove_fixed 1 and at com_maxfps 100 at sv_fps 20
with
pmove_msec 8 pmove_fixed 1 com_maxfps 125 at sv_fps 25
then agian...if its pr default broken... not an option..on a server...
"good old quake.. would run sv_fps 25 ? no problems...right?"
If its "only" wierd flammer graphics, covert ops uni steal time.. and timer pause issues... one could setup a test server..
But if the antilag, antiwarp or anything concering hit detection get messed up.. there is zero piont 2 it...
..i know that the unlagged code becomes less "effective" at high sv_fps..
Still the pmove_msec...hmm... u say 8 is min.. i got it set 3 and 5.. but "if its in effect on server-client"..thats something else? no? i have no idea if i can trust the console printout on that perticular issu..
But if 8 ms is the "must have" choice value and default... the maby one should try sv_fps 25? but im realy not positive .. most likely 99.9999% a complet waist of time...
I'd realy like some feedback on pmove_msec 10 set serverside...
It's a waste of time. SD themselves said stuff breaks when you set sv_fps > 20. It's the nature of the engine, nothing you can do about it (except rewrite the engine parts that are faulty).
Our servers now run on 64 bit steroids. Point your ET to:
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem
Although deus can't jump with b_fixedphysics .ouroboro wrote:b_fixedphysics > pmove_fixed. Maybe not for hardcore trickjumping, but certainly for actually playing the game.
Our servers now run on 64 bit steroids. Point your ET to:
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem
- Forgotten Ground StopWatch Server with occasional wolfrof 1
- Fraggle Rock ETPub Server - Mix up ET/UT & Duke Nukem