<!--quoteo(post=2031086:date=Nov 21 2012, 03:23 PM:name=lwf)--><div class='quotetop'>QUOTE (lwf @ Nov 21 2012, 03:23 PM) <a href="index.php?act=findpost&pid=2031086"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Don't quote me on this because I'm not entirely sure but I believe Max has said the only thing (that we would notice anyway) bound to server tick rate is server controlled entities, like the AI for whips, drifters and MACs.<!--QuoteEnd--></div><!--QuoteEEnd-->
He said exactly that in the Q/A he did a few months ago. But eh, dudes learn how to change a few console variables in CS, so obviously that makes all of them experts in game engines and netcode.
30 tics/sec should be enough from a sampling perspective. Going higher is not going to make any big differences as far as hit-reg goes, but it will lower the latency slightly, reducing the "getting shot behind walls"-effect.
When performance improves, they should remove the clientside cmd limit and then let the servers downsample the input to the tickrate of the server (selected by server owner). (I guess it might already be matching the serverrate automatically...)
<!--quoteo(post=2031103:date=Nov 21 2012, 05:53 AM:name=snaga)--><div class='quotetop'>QUOTE (snaga @ Nov 21 2012, 05:53 AM) <a href="index.php?act=findpost&pid=2031103"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->He said exactly that in the Q/A he did a few months ago. But eh, dudes learn how to change a few console variables in CS, so obviously that makes all of them experts in game engines and netcode.<!--QuoteEnd--></div><!--QuoteEEnd-->
Lol been playing since CS 1.3. I know most every viable cvar of that game better than you know your own mother. Just sayin.
<!--quoteo(post=2031097:date=Nov 21 2012, 05:41 AM:name=Wilson)--><div class='quotetop'>QUOTE (Wilson @ Nov 21 2012, 05:41 AM) <a href="index.php?act=findpost&pid=2031097"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->They developed the engine with lua specifically to support modding. I don't know if they imagined performance would be such a big challenge though. Yet, I'm sure they knew there would be some downsides to using it over C++ but since they came from the modding scene they wanted to create a game that encouraged that. I guess it's debatable whether or not it was worth it.<!--QuoteEnd--></div><!--QuoteEEnd-->
The engine is c++, the game refers to the engine with LUA from what I understand.
<!--quoteo(post=2031196:date=Nov 21 2012, 05:41 PM:name=grizzlyyy)--><div class='quotetop'>QUOTE (grizzlyyy @ Nov 21 2012, 05:41 PM) <a href="index.php?act=findpost&pid=2031196"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Lol been playing since CS 1.3. I know most every viable cvar of that game better than you know your own mother. Just sayin.<!--QuoteEnd--></div><!--QuoteEEnd-->
Doesnt mean that the engine was.. Spark works differently than Source, and knowing cvars is fine and dandy but actually understanding the impacts each has on the netcode is different.. The current cpu usage on servers is already too high for all but the highest peforming cpus or ones that are overclocked, increasing the tickrates would only make that worse.
Comments
He said exactly that in the Q/A he did a few months ago. But eh, dudes learn how to change a few console variables in CS, so obviously that makes all of them experts in game engines and netcode.
When performance improves, they should remove the clientside cmd limit and then let the servers downsample the input to the tickrate of the server (selected by server owner). (I guess it might already be matching the serverrate automatically...)
Lol been playing since CS 1.3. I know most every viable cvar of that game better than you know your own mother. Just sayin.
The engine is c++, the game refers to the engine with LUA from what I understand.
Cool. This is NS2 though.
A game that was inspired by the half-life series.