The client could set a var like "lastattacktime", that gets synced with 
server?

Michael Chang wrote:
> Hi Ryan
> 
> I've moved all the set velocity stuff into gamemovement, using mv->etc
> instead of SetAbsVelocity. This seems to be more reliable in terms of when
> it does things, however I'm still getting a view hitch.
> 
> I've also tried printing out time stamps. An example for net_fakelag 0
> 
> first attack
> client attackstep at 22.305000
> server attackstep at 22.320000
> 
> second attack
> client attackstep at 23.070000
> server attackstep at 23.174999
> 
> With net_fakelag 100
> 
> first attack
> client attackstep at 23.070000
> server attackstep at 23.174999
> 
> second attack
> 
> client attackstep at 249.869995
> server attackstep at 249.974991
> 
> 
> 
> The server falls behind the client often. To my understanding, there's no
> way the server knows when the client player presses the attack button. How
> would one compensate for this major discrepancy?
> 
> 
> ~M
> 
> 
> 
> You might want to write some debug messages for the console with time stamps
> 
> so you can see when the client and server are getting velocities assigned,
> but directly setting the client side velocity is a really bad idea. I agree
> with Jorge that you should be adding your velocity changes directly to the
> shared game movement code. You really need to make sure, and this is really
> important for prediction, that the client and server velocity changes are
> being made in sync with eachother. When the server update hits the client
> and the client is outside of the movement range for that frame, it will snap
> 
> the client back into server sync no matter what.
> _______________________________________________
> 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

Reply via email to