Author Topic: Build 39  (Read 19128 times)

ic3y

  • Committee Member
  • Autococker
  • Posts: 1398
Re: Build 39
« Reply #20 on: October 07, 2013, 12:40:12 PM »
But why is the avg on a local server 58-65 ?
On oXz Private

Ping on Scoreboard -> 21
Lowest -> 17-20 (thats OK)
Avg -> 70-75 (Why?)

yossarianuk

  • VM-68
  • Posts: 127
Re: Build 39
« Reply #21 on: October 07, 2013, 02:33:50 PM »
Linux version works fine !

(Linux sound could still do with some love...)

I like paintball that much I even run a separate Linux distro which still 'supports' OSS (Arch) just for paintball2...

BASEBALLDUDE

  • Autococker
  • Posts: 703
Re: Build 39
« Reply #22 on: October 07, 2013, 04:37:52 PM »
This seems awesome! Can't wait for build 39 to be officially released!

jitspoe

  • Administrator
  • Autococker
  • Posts: 18801
Re: Build 39
« Reply #23 on: October 07, 2013, 05:03:42 PM »
I haven't looked at the code, but I'm guessing average is taking the ping time from the last server update.  The server updates every 100ms (10 times per second).

That means if you have 0 latency and a fast computer and send a packet to the server on the same frame it's doing its update, you could get a 0 ping.  If you send an update just after the server did an update, it won't respond until the next update, meaning your ping would be close to 100ms.  On average, this would be close to 50ms, best case.

I suppose this is accurate, but it's not the numbers we're typically used to.  It will, in practice, take an additional 0-100ms for the server to respond to your input (ex: the time between clicking the mouse button and when a paintball fires).  The average latency is the normal ping + 50ms.

not_payl_obviously

  • 68 Carbine
  • Posts: 415
Re: Build 39
« Reply #24 on: October 07, 2013, 05:43:41 PM »
I have done some quick look on your anticheat Jitspoe, and I must admit I'm surprised with things I found out: new encryption routine (with key), remodeled anti-wallhack and I think hardware ID got some update too? Probably there are more things I haven't been able to look into right now. But surely I'll see if there is something new to change. If you want my advice for future: change "modified content" detection method (you know what you check, and it's ... yhm, lame) and some runtime protection against modification, but still good job on this build (not that much, but progress with almost zero development time). I might write detailed PM later on with details I found on things you have changed and things I have never visited.

As for rest, I haven't tested yet, but when some most annoying bugs will be out I'll check it.

FusSioN

  • Autococker
  • Posts: 603
Re: Build 39
« Reply #25 on: October 07, 2013, 10:50:23 PM »
Yeah, I think I'm contented to b35.
But yeah, True, the autobalance thingy is still in the game

jitspoe

  • Administrator
  • Autococker
  • Posts: 18801
Re: Build 39
« Reply #26 on: October 07, 2013, 11:15:53 PM »
I've found a couple other things I want to fix.  This build will not get an official release.  Hopefully I'll have build 40 out shortly (either tomorrow or next week, since I'm about to go out of town).

As for autobalance, that's a server-side setting: g_autobalance 0 = disabled, g_autobalanced 1 = enabled, allow team switches, g_autobalance 2 = enabled, disable team switches.

jitspoe

  • Administrator
  • Autococker
  • Posts: 18801
Re: Build 39
« Reply #27 on: October 08, 2013, 07:07:49 PM »
I'm curious if anybody has done any performance comparisons to see if build 39 runs faster than build 38.