[hlcoders] Performance : "OB engine" vs "ep1 engine"
Hi, I just finished porting my mod from "ep1 engine" to "OB engine". I wanted to post a quick feedback about performances I had under "ep1 engine" and "OB engine". My mod is based on HL2MP SDK. It is clear that performance under new engine are very bad from what I see: - movement are jittery - fps drops down Even if I set all settings to lowest level nothing is flawless. To ensure it is not due to a bad porting of my mod, I did the same test between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one using "OB engine". I experienced exactly the same performance problem. Do you experienced the same problem ? What can explain such a big difference ? -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
The biggest one would be technological differences. The system requirements are higher than ep1 for starters, because there is a significant technological difference in what is performed. All the shaders are different, and have a lot more computations; Running the hl2mp source in ep1 engine will consume less resources than in orange box, even with the exact same content and code. Take a look at the difference in requirements: HL2/EP1: Minimum: 1.2 GHz Processor, 256MB RAM, DirectX 7 level graphics card, Windows 2000/XP, Mouse, Keyboard, Internet Connection Recommended: 2.4 GHz Processor, 512MB RAM, DirectX 9 level graphics card, Windows 2000/XP, Mouse, Keyboard, Internet Connection EP2/Orange Box: --- Minimum: 1.7 GHz Processor, 512MB RAM, DirectX(R) 8 level Graphics Card, Windows(R) Vista/XP/2000, Mouse, Keyboard, Internet Connection Recommended: Pentium 4 processor (3.0GHz, or better), 1GB RAM, DirectX(R) 9 level Graphics Card, Windows(R) Vista/XP/2000, Mouse, Keyboard, Internet Connection Make sense? :) -Tony On Thu, Apr 24, 2008 at 4:01 AM, Janek <[EMAIL PROTECTED]> wrote: > Hi, > > I just finished porting my mod from "ep1 engine" to "OB engine". I wanted > to > post a quick feedback about performances I had under "ep1 engine" and "OB > engine". My mod is based on HL2MP SDK. > > It is clear that performance under new engine are very bad from what I > see: > - movement are jittery > - fps drops down > Even if I set all settings to lowest level nothing is flawless. > > To ensure it is not due to a bad porting of my mod, I did the same test > between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one > using > "OB engine". I experienced exactly the same performance problem. > > Do you experienced the same problem ? > > What can explain such a big difference ? > > -- > --- > [EMAIL PROTECTED] > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- -Tony ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Somebody with very similar problems fixed them by uninstalling the nVidia Firewall only a few hours ago. Do you have it? Janek wrote: > Hi, > > I just finished porting my mod from "ep1 engine" to "OB engine". I wanted to > post a quick feedback about performances I had under "ep1 engine" and "OB > engine". My mod is based on HL2MP SDK. > > It is clear that performance under new engine are very bad from what I see: > - movement are jittery > - fps drops down > Even if I set all settings to lowest level nothing is flawless. > > To ensure it is not due to a bad porting of my mod, I did the same test > between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one using > "OB engine". I experienced exactly the same performance problem. > > Do you experienced the same problem ? > > What can explain such a big difference ? > > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
It makes sense. I understand that "OB engine" is more power consuming that "ep1 engine" BUT TF2 is running quite good in my computer. In addition my config is : - Athlon 4200X2 - 2GB RAM - 6600GT Why TF2 is running fine but HL2MP SDK is not ? [EMAIL PROTECTED] 2008/4/24 Tony omega Sergi <[EMAIL PROTECTED]>: > The biggest one would be technological differences. The system > requirements > are higher than ep1 for starters, because there is a significant > technological difference in what is performed. > All the shaders are different, and have a lot more computations; Running > the > hl2mp source in ep1 engine will consume less resources than in orange box, > even with the exact same content and code. > > Take a look at the difference in requirements: > > HL2/EP1: > > > Minimum: 1.2 GHz Processor, 256MB RAM, DirectX 7 level graphics card, > Windows 2000/XP, Mouse, Keyboard, Internet Connection > Recommended: 2.4 GHz Processor, 512MB RAM, DirectX 9 level graphics card, > Windows 2000/XP, Mouse, Keyboard, Internet Connection > > > EP2/Orange Box: > --- > Minimum: 1.7 GHz Processor, 512MB RAM, DirectX(R) 8 level Graphics Card, > Windows(R) Vista/XP/2000, Mouse, Keyboard, Internet Connection > Recommended: Pentium 4 processor (3.0GHz, or better), 1GB RAM, DirectX(R) > 9 > level Graphics Card, Windows(R) Vista/XP/2000, Mouse, Keyboard, Internet > Connection > > > Make sense? :) > -Tony > > On Thu, Apr 24, 2008 at 4:01 AM, Janek <[EMAIL PROTECTED]> wrote: > > > Hi, > > > > I just finished porting my mod from "ep1 engine" to "OB engine". I > wanted > > to > > post a quick feedback about performances I had under "ep1 engine" and > "OB > > engine". My mod is based on HL2MP SDK. > > > > It is clear that performance under new engine are very bad from what I > > see: > > - movement are jittery > > - fps drops down > > Even if I set all settings to lowest level nothing is flawless. > > > > To ensure it is not due to a bad porting of my mod, I did the same test > > between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one > > using > > "OB engine". I experienced exactly the same performance problem. > > > > Do you experienced the same problem ? > > > > What can explain such a big difference ? > > > > -- > > --- > > [EMAIL PROTECTED] > > ___ > > To unsubscribe, edit your list preferences, or view the list archives, > > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > > > > > -- > -Tony > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Yeah, I saw this message but I don't have "NVIDIA Forceware Network Access Manager" installed so I don't had to uninstall it. The problem is very different. I really think HL2MP under "OB engine" is haing serious performance problem. [EMAIL PROTECTED] 2008/4/24 Tom Edwards <[EMAIL PROTECTED]>: > Somebody with very similar problems fixed them by uninstalling the > nVidia Firewall only a few hours ago. Do you have it? > > Janek wrote: > > Hi, > > > > I just finished porting my mod from "ep1 engine" to "OB engine". I > wanted to > > post a quick feedback about performances I had under "ep1 engine" and > "OB > > engine". My mod is based on HL2MP SDK. > > > > It is clear that performance under new engine are very bad from what I > see: > > - movement are jittery > > - fps drops down > > Even if I set all settings to lowest level nothing is flawless. > > > > To ensure it is not due to a bad porting of my mod, I did the same test > > between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one > using > > "OB engine". I experienced exactly the same performance problem. > > > > Do you experienced the same problem ? > > > > What can explain such a big difference ? > > > > > > > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
There is a "Stuttering" problem with it (Affects movement). It seems to lag with this. Janek wrote: > Yeah, I saw this message but I don't have "NVIDIA Forceware Network Access > Manager" installed so I don't had to uninstall it. The problem is very > different. I really think HL2MP under "OB engine" is haing serious > performance problem. > > [EMAIL PROTECTED] > > 2008/4/24 Tom Edwards <[EMAIL PROTECTED]>: > > >> Somebody with very similar problems fixed them by uninstalling the >> nVidia Firewall only a few hours ago. Do you have it? >> >> Janek wrote: >> >>> Hi, >>> >>> I just finished porting my mod from "ep1 engine" to "OB engine". I >>> >> wanted to >> >>> post a quick feedback about performances I had under "ep1 engine" and >>> >> "OB >> >>> engine". My mod is based on HL2MP SDK. >>> >>> It is clear that performance under new engine are very bad from what I >>> >> see: >> >>> - movement are jittery >>> - fps drops down >>> Even if I set all settings to lowest level nothing is flawless. >>> >>> To ensure it is not due to a bad porting of my mod, I did the same test >>> between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one >>> >> using >> >>> "OB engine". I experienced exactly the same performance problem. >>> >>> Do you experienced the same problem ? >>> >>> What can explain such a big difference ? >>> >>> >>> >> ___ >> To unsubscribe, edit your list preferences, or view the list archives, >> please visit: >> http://list.valvesoftware.com/mailman/listinfo/hlcoders >> >> >> > > > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
If you're referring to player movement being laggy and prediction going out, it's because in the released source right now the network tables aren't setup correctly for hl2mp or the scratch sdk. They're missing the new nonlocal data table, and local data overrides which setup the network origin for multiplayer using optimized values. - full precision is sent to the local client, and every other player gets an optimized version that isn't as precise ( interp on the client eliminates it visually ). Whenever we get the SDK update again, that issue will be fixed properly. I've corrected it, but no ETA. -Tony On Thu, Apr 24, 2008 at 5:01 AM, Tom Leighton <[EMAIL PROTECTED]> wrote: > There is a "Stuttering" problem with it (Affects movement). It seems to > lag with this. > > Janek wrote: > > Yeah, I saw this message but I don't have "NVIDIA Forceware Network > Access > > Manager" installed so I don't had to uninstall it. The problem is very > > different. I really think HL2MP under "OB engine" is haing serious > > performance problem. > > > > [EMAIL PROTECTED] > > > > 2008/4/24 Tom Edwards <[EMAIL PROTECTED]>: > > > > > >> Somebody with very similar problems fixed them by uninstalling the > >> nVidia Firewall only a few hours ago. Do you have it? > >> > >> Janek wrote: > >> > >>> Hi, > >>> > >>> I just finished porting my mod from "ep1 engine" to "OB engine". I > >>> > >> wanted to > >> > >>> post a quick feedback about performances I had under "ep1 engine" and > >>> > >> "OB > >> > >>> engine". My mod is based on HL2MP SDK. > >>> > >>> It is clear that performance under new engine are very bad from what I > >>> > >> see: > >> > >>> - movement are jittery > >>> - fps drops down > >>> Even if I set all settings to lowest level nothing is flawless. > >>> > >>> To ensure it is not due to a bad porting of my mod, I did the same > test > >>> between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one > >>> > >> using > >> > >>> "OB engine". I experienced exactly the same performance problem. > >>> > >>> Do you experienced the same problem ? > >>> > >>> What can explain such a big difference ? > >>> > >>> > >>> > >> ___ > >> To unsubscribe, edit your list preferences, or view the list archives, > >> please visit: > >> http://list.valvesoftware.com/mailman/listinfo/hlcoders > >> > >> > >> > > > > > > > > > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- -Tony ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
I was also thinking on a prediction error reading an other thread speaking about that. Thank you for the information Tony. Sad fix is not published yet. 2008/4/24 Tony omega Sergi <[EMAIL PROTECTED]>: > If you're referring to player movement being laggy and prediction going > out, > it's because in the released source right now the network tables aren't > setup correctly for hl2mp or the scratch sdk. > They're missing the new nonlocal data table, and local data overrides which > setup the network origin for multiplayer using optimized values. - full > precision is sent to the local client, and every other player gets an > optimized version that isn't as precise ( interp on the client eliminates > it > visually ). > > Whenever we get the SDK update again, that issue will be fixed properly. > I've corrected it, but no ETA. > -Tony > > On Thu, Apr 24, 2008 at 5:01 AM, Tom Leighton <[EMAIL PROTECTED] > > > wrote: > > > There is a "Stuttering" problem with it (Affects movement). It seems to > > lag with this. > > > > Janek wrote: > > > Yeah, I saw this message but I don't have "NVIDIA Forceware Network > > Access > > > Manager" installed so I don't had to uninstall it. The problem is very > > > different. I really think HL2MP under "OB engine" is haing serious > > > performance problem. > > > > > > [EMAIL PROTECTED] > > > > > > 2008/4/24 Tom Edwards <[EMAIL PROTECTED]>: > > > > > > > > >> Somebody with very similar problems fixed them by uninstalling the > > >> nVidia Firewall only a few hours ago. Do you have it? > > >> > > >> Janek wrote: > > >> > > >>> Hi, > > >>> > > >>> I just finished porting my mod from "ep1 engine" to "OB engine". I > > >>> > > >> wanted to > > >> > > >>> post a quick feedback about performances I had under "ep1 engine" and > > >>> > > >> "OB > > >> > > >>> engine". My mod is based on HL2MP SDK. > > >>> > > >>> It is clear that performance under new engine are very bad from what > I > > >>> > > >> see: > > >> > > >>> - movement are jittery > > >>> - fps drops down > > >>> Even if I set all settings to lowest level nothing is flawless. > > >>> > > >>> To ensure it is not due to a bad porting of my mod, I did the same > > test > > >>> between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the one > > >>> > > >> using > > >> > > >>> "OB engine". I experienced exactly the same performance problem. > > >>> > > >>> Do you experienced the same problem ? > > >>> > > >>> What can explain such a big difference ? > > >>> > > >>> > > >>> > > >> ___ > > >> To unsubscribe, edit your list preferences, or view the list archives, > > >> please visit: > > >> http://list.valvesoftware.com/mailman/listinfo/hlcoders > > >> > > >> > > >> > > > > > > > > > > > > > > > ___ > > To unsubscribe, edit your list preferences, or view the list archives, > > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > > > > > -- > -Tony > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Tony, you could maybe post what is missing in those tables. :-D 2008/4/24 Janek <[EMAIL PROTECTED]>: > I was also thinking on a prediction error reading an other thread speaking > about that. Thank you for the information Tony. > Sad fix is not published yet. > > 2008/4/24 Tony omega Sergi <[EMAIL PROTECTED]>: > >> If you're referring to player movement being laggy and prediction going >> out, >> >> it's because in the released source right now the network tables aren't >> setup correctly for hl2mp or the scratch sdk. >> They're missing the new nonlocal data table, and local data overrides >> which >> setup the network origin for multiplayer using optimized values. - full >> precision is sent to the local client, and every other player gets an >> optimized version that isn't as precise ( interp on the client eliminates >> it >> visually ). >> >> Whenever we get the SDK update again, that issue will be fixed properly. >> I've corrected it, but no ETA. >> -Tony >> >> On Thu, Apr 24, 2008 at 5:01 AM, Tom Leighton < >> [EMAIL PROTECTED]> >> wrote: >> >> > There is a "Stuttering" problem with it (Affects movement). It seems to >> > lag with this. >> > >> > Janek wrote: >> > > Yeah, I saw this message but I don't have "NVIDIA Forceware Network >> > Access >> > > Manager" installed so I don't had to uninstall it. The problem is very >> > > different. I really think HL2MP under "OB engine" is haing serious >> > > performance problem. >> > > >> > > [EMAIL PROTECTED] >> > > >> > > 2008/4/24 Tom Edwards <[EMAIL PROTECTED]>: >> > > >> > > >> > >> Somebody with very similar problems fixed them by uninstalling the >> > >> nVidia Firewall only a few hours ago. Do you have it? >> > >> >> > >> Janek wrote: >> > >> >> > >>> Hi, >> > >>> >> > >>> I just finished porting my mod from "ep1 engine" to "OB engine". I >> > >>> >> > >> wanted to >> > >> >> > >>> post a quick feedback about performances I had under "ep1 engine" >> and >> > >>> >> > >> "OB >> > >> >> > >>> engine". My mod is based on HL2MP SDK. >> > >>> >> > >>> It is clear that performance under new engine are very bad from what >> I >> > >>> >> > >> see: >> > >> >> > >>> - movement are jittery >> > >>> - fps drops down >> > >>> Even if I set all settings to lowest level nothing is flawless. >> > >>> >> > >>> To ensure it is not due to a bad porting of my mod, I did the same >> > test >> > >>> between a fresh HL2MP SDK (from Valve) using "ep1 engine" and the >> one >> > >>> >> > >> using >> > >> >> > >>> "OB engine". I experienced exactly the same performance problem. >> > >>> >> > >>> Do you experienced the same problem ? >> > >>> >> > >>> What can explain such a big difference ? >> > >>> >> > >>> >> > >>> >> > >> ___ >> > >> To unsubscribe, edit your list preferences, or view the list >> archives, >> > >> please visit: >> > >> http://list.valvesoftware.com/mailman/listinfo/hlcoders >> > >> >> > >> >> > >> >> > > >> > > >> > > >> > >> > >> > ___ >> > To unsubscribe, edit your list preferences, or view the list archives, >> > please visit: >> > http://list.valvesoftware.com/mailman/listinfo/hlcoders >> > >> > >> >> >> -- >> -Tony >> ___ >> To unsubscribe, edit your list preferences, or view the list archives, >> please visit: >> http://list.valvesoftware.com/mailman/listinfo/hlcoders >> >> > > > -- > --- > [EMAIL PROTECTED] -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Here are the ones for sdkplayer:
note that when the sdk gets updated these will change again, but this should
get you started for now - you can make the same changes to hl2mp_player as
necessary.
sdk_player.cpp:
---
//-
// Purpose: Filters updates to a variable so that only non-local players see
// the changes. This is so we can send a low-res origin to non-local
players
// while sending a hi-res one to the local player.
// Input : *pVarData -
//*pOut -
//objectID -
//-
void* SendProxy_SendNonLocalDataTable( const SendProp *pProp, const void
*pStruct, const void *pVarData, CSendProxyRecipients *pRecipients, int
objectID )
{
pRecipients->SetAllRecipients();
pRecipients->ClearRecipient( objectID - 1 );
return ( void * )pVarData;
}
REGISTER_SEND_PROXY_NON_MODIFIED_POINTER( SendProxy_SendNonLocalDataTable );
BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, DT_SDKLocalPlayerExclusive )
SendPropInt( SENDINFO( m_iShotsFired ), 8, SPROP_UNSIGNED ),
// send a hi-res origin to the local player for use in prediction
SendPropVector(SENDINFO(m_vecOrigin), -1,
SPROP_NOSCALE|SPROP_CHANGES_OFTEN, 0.0f, HIGH_DEFAULT, SendProxy_Origin ),
SendPropFloat( SENDINFO_VECTORELEM(m_angEyeAngles, 0), 8,
SPROP_CHANGES_OFTEN, -90.0f, 90.0f ),
//SendPropAngle( SENDINFO_VECTORELEM(m_angEyeAngles, 1), 10,
SPROP_CHANGES_OFTEN ),
END_SEND_TABLE()
BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, DT_SDKNonLocalPlayerExclusive )
// send a lo-res origin to other players
SendPropVector(SENDINFO(m_vecOrigin), -1,
SPROP_COORD_MP_LOWPRECISION|SPROP_CHANGES_OFTEN, 0.0f, HIGH_DEFAULT,
SendProxy_Origin ),
SendPropFloat( SENDINFO_VECTORELEM(m_angEyeAngles, 0), 8,
SPROP_CHANGES_OFTEN, -90.0f, 90.0f ),
SendPropAngle( SENDINFO_VECTORELEM(m_angEyeAngles, 1), 10,
SPROP_CHANGES_OFTEN ),
END_SEND_TABLE()
// main table
IMPLEMENT_SERVERCLASS_ST( CSDKPlayer, DT_SDKPlayer )
SendPropExclude( "DT_BaseAnimating", "m_flPoseParameter" ),
SendPropExclude( "DT_BaseAnimating", "m_flPlaybackRate" ),
SendPropExclude( "DT_BaseAnimating", "m_nSequence" ),
SendPropExclude( "DT_BaseEntity", "m_angRotation" ),
SendPropExclude( "DT_BaseAnimatingOverlay", "overlay_vars" ),
SendPropExclude( "DT_BaseEntity", "m_vecOrigin" ),
// playeranimstate and clientside animation takes care of these on the
client
SendPropExclude( "DT_ServerAnimationData" , "m_flCycle" ),
SendPropExclude( "DT_AnimTimeMustBeFirst" , "m_flAnimTime" ),
// Data that only gets sent to the local player.
SendPropDataTable( "sdklocaldata", 0,
&REFERENCE_SEND_TABLE(DT_SDKLocalPlayerExclusive),
SendProxy_SendLocalDataTable ),
// Data that gets sent to all other players
SendPropDataTable( "sdknonlocaldata", 0,
&REFERENCE_SEND_TABLE(DT_SDKNonLocalPlayerExclusive),
SendProxy_SendNonLocalDataTable ),
SendPropEHandle( SENDINFO( m_hRagdoll ) ),
SendPropInt( SENDINFO( m_iPlayerState ), Q_log2( NUM_PLAYER_STATES )+1,
SPROP_UNSIGNED ),
SendPropBool( SENDINFO( m_bSpawnInterpCounter ) ),
END_SEND_TABLE()
c_sdk_player.cpp:
--
BEGIN_RECV_TABLE_NOBASE( C_SDKPlayer, DT_SDKLocalPlayerExclusive )
RecvPropInt( RECVINFO( m_iShotsFired ) ),
RecvPropVector( RECVINFO_NAME( m_vecNetworkOrigin, m_vecOrigin ) ),
RecvPropFloat( RECVINFO( m_angEyeAngles[0] ) ),
//RecvPropFloat( RECVINFO( m_angEyeAngles[1] ) ),
END_RECV_TABLE()
BEGIN_RECV_TABLE_NOBASE( C_SDKPlayer, DT_SDKNonLocalPlayerExclusive )
RecvPropVector( RECVINFO_NAME( m_vecNetworkOrigin, m_vecOrigin ) ),
RecvPropFloat( RECVINFO( m_angEyeAngles[0] ) ),
RecvPropFloat( RECVINFO( m_angEyeAngles[1] ) ),
END_RECV_TABLE()
// main table
IMPLEMENT_CLIENTCLASS_DT( C_SDKPlayer, DT_SDKPlayer, CSDKPlayer )
RecvPropDataTable( "sdklocaldata", 0, 0,
&REFERENCE_RECV_TABLE(DT_SDKLocalPlayerExclusive) ),
RecvPropDataTable( "sdknonlocaldata", 0, 0,
&REFERENCE_RECV_TABLE(DT_SDKNonLocalPlayerExclusive) ),
RecvPropEHandle( RECVINFO( m_hRagdoll ) ),
RecvPropInt( RECVINFO( m_iPlayerState ) ),
RecvPropBool( RECVINFO( m_bSpawnInterpCounter ) ),
END_RECV_TABLE()
On Thu, Apr 24, 2008 at 5:41 PM, Janek <[EMAIL PROTECTED]> wrote:
> Tony, you could maybe post what is missing in those tables. :-D
>
> 2008/4/24 Janek <[EMAIL PROTECTED]>:
>
> > I was also thinking on a prediction error reading an other thread
> speaking
> > about that. Thank you for the information Tony.
> > Sad fix is not published yet.
> >
> > 2008/4/24 Tony omega Sergi <[EMAIL PROTECTED]>:
> >
> >> If you're referring to player movement being laggy and prediction going
> >> out,
> >>
> >> it's because in the released source right now the network tables aren't
> >> setup correctly for hl2mp or the scratch sdk.
> >> They're missing the new nonlocal data table, and local
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
I am really shocked about this massive performance difference. I read
about this nowhere else, I had no idea this would happen. I am only
very thankful that I didn't bother upgrading to OB sdk. Does anyone
know if/when these performance issues will be resolved
On Thu, Apr 24, 2008 at 5:00 PM, Tony omega Sergi <[EMAIL PROTECTED]> wrote:
> Here are the ones for sdkplayer:
> note that when the sdk gets updated these will change again, but this should
> get you started for now - you can make the same changes to hl2mp_player as
> necessary.
>
> sdk_player.cpp:
> ---
>
>
> //-
> // Purpose: Filters updates to a variable so that only non-local players see
> // the changes. This is so we can send a low-res origin to non-local
> players
> // while sending a hi-res one to the local player.
> // Input : *pVarData -
> //*pOut -
> //objectID -
>
> //-
>
> void* SendProxy_SendNonLocalDataTable( const SendProp *pProp, const void
> *pStruct, const void *pVarData, CSendProxyRecipients *pRecipients, int
> objectID )
> {
> pRecipients->SetAllRecipients();
> pRecipients->ClearRecipient( objectID - 1 );
> return ( void * )pVarData;
> }
> REGISTER_SEND_PROXY_NON_MODIFIED_POINTER( SendProxy_SendNonLocalDataTable );
>
> BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, DT_SDKLocalPlayerExclusive )
> SendPropInt( SENDINFO( m_iShotsFired ), 8, SPROP_UNSIGNED ),
> // send a hi-res origin to the local player for use in prediction
> SendPropVector(SENDINFO(m_vecOrigin), -1,
> SPROP_NOSCALE|SPROP_CHANGES_OFTEN, 0.0f, HIGH_DEFAULT, SendProxy_Origin ),
>
> SendPropFloat( SENDINFO_VECTORELEM(m_angEyeAngles, 0), 8,
> SPROP_CHANGES_OFTEN, -90.0f, 90.0f ),
> //SendPropAngle( SENDINFO_VECTORELEM(m_angEyeAngles, 1), 10,
> SPROP_CHANGES_OFTEN ),
> END_SEND_TABLE()
>
> BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, DT_SDKNonLocalPlayerExclusive )
> // send a lo-res origin to other players
> SendPropVector(SENDINFO(m_vecOrigin), -1,
> SPROP_COORD_MP_LOWPRECISION|SPROP_CHANGES_OFTEN, 0.0f, HIGH_DEFAULT,
> SendProxy_Origin ),
>
> SendPropFloat( SENDINFO_VECTORELEM(m_angEyeAngles, 0), 8,
> SPROP_CHANGES_OFTEN, -90.0f, 90.0f ),
> SendPropAngle( SENDINFO_VECTORELEM(m_angEyeAngles, 1), 10,
> SPROP_CHANGES_OFTEN ),
> END_SEND_TABLE()
>
>
> // main table
> IMPLEMENT_SERVERCLASS_ST( CSDKPlayer, DT_SDKPlayer )
> SendPropExclude( "DT_BaseAnimating", "m_flPoseParameter" ),
> SendPropExclude( "DT_BaseAnimating", "m_flPlaybackRate" ),
> SendPropExclude( "DT_BaseAnimating", "m_nSequence" ),
> SendPropExclude( "DT_BaseEntity", "m_angRotation" ),
> SendPropExclude( "DT_BaseAnimatingOverlay", "overlay_vars" ),
> SendPropExclude( "DT_BaseEntity", "m_vecOrigin" ),
>
> // playeranimstate and clientside animation takes care of these on the
> client
> SendPropExclude( "DT_ServerAnimationData" , "m_flCycle" ),
> SendPropExclude( "DT_AnimTimeMustBeFirst" , "m_flAnimTime" ),
>
> // Data that only gets sent to the local player.
> SendPropDataTable( "sdklocaldata", 0,
> &REFERENCE_SEND_TABLE(DT_SDKLocalPlayerExclusive),
> SendProxy_SendLocalDataTable ),
> // Data that gets sent to all other players
> SendPropDataTable( "sdknonlocaldata", 0,
> &REFERENCE_SEND_TABLE(DT_SDKNonLocalPlayerExclusive),
> SendProxy_SendNonLocalDataTable ),
>
> SendPropEHandle( SENDINFO( m_hRagdoll ) ),
>
> SendPropInt( SENDINFO( m_iPlayerState ), Q_log2( NUM_PLAYER_STATES )+1,
> SPROP_UNSIGNED ),
>
> SendPropBool( SENDINFO( m_bSpawnInterpCounter ) ),
>
> END_SEND_TABLE()
>
> c_sdk_player.cpp:
> --
> BEGIN_RECV_TABLE_NOBASE( C_SDKPlayer, DT_SDKLocalPlayerExclusive )
> RecvPropInt( RECVINFO( m_iShotsFired ) ),
> RecvPropVector( RECVINFO_NAME( m_vecNetworkOrigin, m_vecOrigin ) ),
>
> RecvPropFloat( RECVINFO( m_angEyeAngles[0] ) ),
> //RecvPropFloat( RECVINFO( m_angEyeAngles[1] ) ),
> END_RECV_TABLE()
>
> BEGIN_RECV_TABLE_NOBASE( C_SDKPlayer, DT_SDKNonLocalPlayerExclusive )
> RecvPropVector( RECVINFO_NAME( m_vecNetworkOrigin, m_vecOrigin ) ),
>
> RecvPropFloat( RECVINFO( m_angEyeAngles[0] ) ),
> RecvPropFloat( RECVINFO( m_angEyeAngles[1] ) ),
> END_RECV_TABLE()
>
> // main table
> IMPLEMENT_CLIENTCLASS_DT( C_SDKPlayer, DT_SDKPlayer, CSDKPlayer )
> RecvPropDataTable( "sdklocaldata", 0, 0,
> &REFERENCE_RECV_TABLE(DT_SDKLocalPlayerExclusive) ),
> RecvPropDataTable( "sdknonlocaldata", 0, 0,
> &REFERENCE_RECV_TABLE(DT_SDKNonLocalPlayerExclusive) ),
>
> RecvPropEHandle( RECVINFO( m_hRagdoll ) ),
>
> RecvPropInt( RECVINFO( m_iPlayerState ) ),
>
> RecvPropBool( RECVINFO( m_bSpawnInterpCounter ) ),
> END_RECV_TABLE()
>
>
>
>
>
> On Thu, Apr 24, 2008 at 5:41 PM, Janek <[EMAIL PROTECTED]>
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Maybe when some severe technical malfunction causes all computers in the world to be downgraded to technology from 2005? :) -Tony On Fri, Apr 25, 2008 at 12:51 AM, Nick <[EMAIL PROTECTED]> wrote: > I am really shocked about this massive performance difference. I read > about this nowhere else, I had no idea this would happen. I am only > very thankful that I didn't bother upgrading to OB sdk. Does anyone > know if/when these performance issues will be resolved > > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
As it is slightly different in HL2MP SDK and as there is still no linux version of SDK I will wait next update of Valve. It is probably the best things to do to avoid back changes. Anyway, thank you very much for the information Tony. [EMAIL PROTECTED] 2008/4/25 Tony omega Sergi <[EMAIL PROTECTED]>: > Maybe when some severe technical malfunction causes all computers in the > world to be downgraded to technology from 2005? :) > -Tony > > On Fri, Apr 25, 2008 at 12:51 AM, Nick <[EMAIL PROTECTED]> wrote: > > > I am really shocked about this massive performance difference. I read > > about this nowhere else, I had no idea this would happen. I am only > > very thankful that I didn't bother upgrading to OB sdk. Does anyone > > know if/when these performance issues will be resolved > > > > > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Nick wrote: > I am really shocked about this massive performance difference. I read > about this nowhere else, I had no idea this would happen. I am only > very thankful that I didn't bother upgrading to OB sdk. Does anyone > know if/when these performance issues will be resolved Probably never. Next-gen games require next-gen hardware. You're not going to be able to run more recent games with your '386 computer and your GeForce2 graphics card. :) Seriously, game developers will keep pushing the top-end hardware just as far as it can go. If you don't keep upgrading your hardware every few years, you will be unable to run the latest and greatest games. -- Jeffrey "botman" Broome ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
That reply only makes sense if the OB engine has some serious features added that require these performances. If you want the same functionality on a more modern engine, the performance should stay the same or increase, not decrease. I don't know what they added in the OB engine that justifies a performance drop ( didn't even know there was one ), I'm perfectly happy with my ep1 functionality & performance, just saying that you shouldn't take for granted that developers need a bigger hardware budget each time unless it's justified. Not to mention they're supporting a console which has at least a 5 year lifespan with a fixed performance budget. > Nick wrote: >> I am really shocked about this massive performance difference. I read >> about this nowhere else, I had no idea this would happen. I am only >> very thankful that I didn't bother upgrading to OB sdk. Does anyone >> know if/when these performance issues will be resolved > > Probably never. Next-gen games require next-gen hardware. You're not > going to be able to run more recent games with your '386 computer and > your GeForce2 graphics card. :) > > Seriously, game developers will keep pushing the top-end hardware just > as far as it can go. If you don't keep upgrading your hardware every > few years, you will be unable to run the latest and greatest games. > > -- > Jeffrey "botman" Broome > > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
I'd say the dynamic shadows, and the new particle system are the causes. /ScarT On 25/04/2008, Maarten De Meyer <[EMAIL PROTECTED]> wrote: > > That reply only makes sense if the OB engine has some serious features > added that require these performances. If you want the same functionality > on a more modern engine, the performance should stay the same or increase, > not decrease. I don't know what they added in the OB engine that justifies > a performance drop ( didn't even know there was one ), I'm perfectly happy > with my ep1 functionality & performance, just saying that you shouldn't > take for granted that developers need a bigger hardware budget each time > unless it's justified. > > Not to mention they're supporting a console which has at least a 5 year > lifespan with a fixed performance budget. > > > Nick wrote: > >> I am really shocked about this massive performance difference. I read > >> about this nowhere else, I had no idea this would happen. I am only > >> very thankful that I didn't bother upgrading to OB sdk. Does anyone > >> know if/when these performance issues will be resolved > > > > Probably never. Next-gen games require next-gen hardware. You're not > > going to be able to run more recent games with your '386 computer and > > your GeForce2 graphics card. :) > > > > Seriously, game developers will keep pushing the top-end hardware just > > as far as it can go. If you don't keep upgrading your hardware every > > few years, you will be unable to run the latest and greatest games. > > > > -- > > Jeffrey "botman" Broome > > > > ___ > > To unsubscribe, edit your list preferences, or view the list archives, > > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > > > > > > > > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
I blame tony's terrible coding. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Just to clarify. The problem I was refering to is not a real performance issue. The problem is only related to laggy player movement due to prediction issue. TF2 is working quite well in my 6600GT so "OB engine" is not requiring so much additional power. 2008/4/25 Andrew Ritchie <[EMAIL PROTECTED]>: > I blame tony's terrible coding. > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
Tony, There are things which make me not understanding how your code can fix the jittery movement issue. I try to explain hereafter. 1) I understand you add "nonlocaldata" tables. This table wasn't present in previous version (ep1 engine). I understand it is sending variables for other players (not local player). It will generate additional network flow. I 'm not sure it is so important. Also my problem related to jittery movement is linked to local player. Why do we really need it ? 2) You send "m_angEyeAngles" in "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, DT_SDKLocalPlayerExclusive )" and "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, DT_SDKLocalPlayerExclusive )" even if it is already present in the original codeline in "IMPLEMENT_SERVERCLASS_ST( CSDKPlayer, DT_SDKPlayer )" What does it really change ? 3) You send "m_vecOrigin" It is already send in "IMPLEMENT_SERVERCLASS_ST_NOBASE( CBaseEntity, DT_BaseEntity )" (baseentity.cpp) Before implementing any solution in my codeline, I would like to understand coz atm it is a mystery for me. J. 2008/4/25 Janek <[EMAIL PROTECTED]>: > Just to clarify. The problem I was refering to is not a real performance > issue. The problem is only related to laggy player movement due to > prediction issue. > TF2 is working quite well in my 6600GT so "OB engine" is not requiring so > much additional power. > > 2008/4/25 Andrew Ritchie <[EMAIL PROTECTED]>: > >> I blame tony's terrible coding. >> ___ >> To unsubscribe, edit your list preferences, or view the list archives, >> please visit: >> http://list.valvesoftware.com/mailman/listinfo/hlcoders >> >> > > > -- > --- > [EMAIL PROTECTED] -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
I think j is talking about the jittery viewmodels, as that is also an issue for us too. 2008/6/16 Janek <[EMAIL PROTECTED]>: > Tony, > > There are things which make me not understanding how your code can fix the > jittery movement issue. I try to explain hereafter. > > 1) I understand you add "nonlocaldata" tables. > This table wasn't present in previous version (ep1 engine). I understand it > is sending variables for other players (not local player). It will generate > additional network flow. I 'm not sure it is so important. Also my problem > related to jittery movement is linked to local player. Why do we really > need > it ? > > 2) You send "m_angEyeAngles" in "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, > DT_SDKLocalPlayerExclusive )" and "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, > DT_SDKLocalPlayerExclusive )" even if it is already present in the original > codeline in "IMPLEMENT_SERVERCLASS_ST( CSDKPlayer, DT_SDKPlayer )" > What does it really change ? > > 3) You send "m_vecOrigin" > It is already send in "IMPLEMENT_SERVERCLASS_ST_NOBASE( CBaseEntity, > DT_BaseEntity )" (baseentity.cpp) > Before implementing any solution in my codeline, I would like to understand > coz atm it is a mystery for me. > > J. > 2008/4/25 Janek <[EMAIL PROTECTED]>: > > > Just to clarify. The problem I was refering to is not a real performance > > issue. The problem is only related to laggy player movement due to > > prediction issue. > > TF2 is working quite well in my 6600GT so "OB engine" is not requiring so > > much additional power. > > > > 2008/4/25 Andrew Ritchie <[EMAIL PROTECTED]>: > > > >> I blame tony's terrible coding. > >> ___ > >> To unsubscribe, edit your list preferences, or view the list archives, > >> please visit: > >> http://list.valvesoftware.com/mailman/listinfo/hlcoders > >> > >> > > > > > > -- > > --- > > [EMAIL PROTECTED] > > > > > -- > --- > [EMAIL PROTECTED] > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
simple. the base code has the origin transmission optimized so it sends less data over the pipe, because interpolation will take care of all of that and not be noticeable. but, the local player doesn't interpolate, it predicts and maintains syncronization, it needs to be accurate. those table changes make it so the local player gets full data like ep1, but makes it so remote players don't send you high res data, since your client can interpolate the changes anyway. so now the local player is not jittering (it's the prediction getting stomped on, joel (oneeyed) posted about this months ago, and this fixes it) and the networking stays optimized as much as possible. On Mon, Jun 16, 2008 at 10:04 AM, Olly <[EMAIL PROTECTED]> wrote: > I think j is talking about the jittery viewmodels, as that is also an issue > for us too. > > 2008/6/16 Janek <[EMAIL PROTECTED]>: > > > Tony, > > > > There are things which make me not understanding how your code can fix > the > > jittery movement issue. I try to explain hereafter. > > > > 1) I understand you add "nonlocaldata" tables. > > This table wasn't present in previous version (ep1 engine). I understand > it > > is sending variables for other players (not local player). It will > generate > > additional network flow. I 'm not sure it is so important. Also my > problem > > related to jittery movement is linked to local player. Why do we really > > need > > it ? > > > > 2) You send "m_angEyeAngles" in "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, > > DT_SDKLocalPlayerExclusive )" and "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, > > DT_SDKLocalPlayerExclusive )" even if it is already present in the > original > > codeline in "IMPLEMENT_SERVERCLASS_ST( CSDKPlayer, DT_SDKPlayer )" > > What does it really change ? > > > > 3) You send "m_vecOrigin" > > It is already send in "IMPLEMENT_SERVERCLASS_ST_NOBASE( CBaseEntity, > > DT_BaseEntity )" (baseentity.cpp) > > Before implementing any solution in my codeline, I would like to > understand > > coz atm it is a mystery for me. > > > > J. > > 2008/4/25 Janek <[EMAIL PROTECTED]>: > > > > > Just to clarify. The problem I was refering to is not a real > performance > > > issue. The problem is only related to laggy player movement due to > > > prediction issue. > > > TF2 is working quite well in my 6600GT so "OB engine" is not requiring > so > > > much additional power. > > > > > > 2008/4/25 Andrew Ritchie <[EMAIL PROTECTED]>: > > > > > >> I blame tony's terrible coding. > > >> ___ > > >> To unsubscribe, edit your list preferences, or view the list archives, > > >> please visit: > > >> http://list.valvesoftware.com/mailman/listinfo/hlcoders > > >> > > >> > > > > > > > > > -- > > > --- > > > [EMAIL PROTECTED] > > > > > > > > > > -- > > --- > > [EMAIL PROTECTED] > > ___ > > To unsubscribe, edit your list preferences, or view the list archives, > > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > > > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- -Tony ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
Re: [hlcoders] Performance : "OB engine" vs "ep1 engine"
OK. I'll try to implement this solution taking care of hl2mp context and test as soon as I have time to. Thank you for your explainations. J. 2008/6/16 Tony omega Sergi <[EMAIL PROTECTED]>: > simple. > > the base code has the origin transmission optimized so it sends less data > over the pipe, because interpolation will take care of all of that and not > be noticeable. but, the local player doesn't interpolate, it predicts and > maintains syncronization, it needs to be accurate. > > those table changes make it so the local player gets full data like ep1, > but > makes it so remote players don't send you high res data, since your client > can interpolate the changes anyway. > > so now the local player is not jittering (it's the prediction getting > stomped on, joel (oneeyed) posted about this months ago, and this fixes it) > and the networking stays optimized as much as possible. > > > On Mon, Jun 16, 2008 at 10:04 AM, Olly <[EMAIL PROTECTED]> wrote: > > > I think j is talking about the jittery viewmodels, as that is also an > issue > > for us too. > > > > 2008/6/16 Janek <[EMAIL PROTECTED]>: > > > > > Tony, > > > > > > There are things which make me not understanding how your code can fix > > the > > > jittery movement issue. I try to explain hereafter. > > > > > > 1) I understand you add "nonlocaldata" tables. > > > This table wasn't present in previous version (ep1 engine). I > understand > > it > > > is sending variables for other players (not local player). It will > > generate > > > additional network flow. I 'm not sure it is so important. Also my > > problem > > > related to jittery movement is linked to local player. Why do we really > > > need > > > it ? > > > > > > 2) You send "m_angEyeAngles" in "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, > > > DT_SDKLocalPlayerExclusive )" and "BEGIN_SEND_TABLE_NOBASE( CSDKPlayer, > > > DT_SDKLocalPlayerExclusive )" even if it is already present in the > > original > > > codeline in "IMPLEMENT_SERVERCLASS_ST( CSDKPlayer, DT_SDKPlayer )" > > > What does it really change ? > > > > > > 3) You send "m_vecOrigin" > > > It is already send in "IMPLEMENT_SERVERCLASS_ST_NOBASE( CBaseEntity, > > > DT_BaseEntity )" (baseentity.cpp) > > > Before implementing any solution in my codeline, I would like to > > understand > > > coz atm it is a mystery for me. > > > > > > J. > > > 2008/4/25 Janek <[EMAIL PROTECTED]>: > > > > > > > Just to clarify. The problem I was refering to is not a real > > performance > > > > issue. The problem is only related to laggy player movement due to > > > > prediction issue. > > > > TF2 is working quite well in my 6600GT so "OB engine" is not > requiring > > so > > > > much additional power. > > > > > > > > 2008/4/25 Andrew Ritchie <[EMAIL PROTECTED]>: > > > > > > > >> I blame tony's terrible coding. > > > >> ___ > > > >> To unsubscribe, edit your list preferences, or view the list > archives, > > > >> please visit: > > > >> http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > >> > > > >> > > > > > > > > > > > > -- > > > > --- > > > > [EMAIL PROTECTED] > > > > > > > > > > > > > > > -- > > > --- > > > [EMAIL PROTECTED] > > > ___ > > > To unsubscribe, edit your list preferences, or view the list archives, > > > please visit: > > > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > > > > > > ___ > > To unsubscribe, edit your list preferences, or view the list archives, > > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > > > > > > -- > -Tony > ___ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlcoders > > -- --- [EMAIL PROTECTED] ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlcoders
