[Xoleum]

Originally Posted by usla Will wait for CSS 2 then )

That does sound like... nothing accomplished?

[rhelgeby]

http://code.google.com/p/zombiereloa...s/detail?id=16

See comment 3 by me in that issue. There's still hope.

[Xoleum]

Originally Posted by rhelgeby http://code.google.com/p/zombiereloa...s/detail?id=16 See comment 3 by me in that issue. There's still hope.

Awesome! More info please! =D

[KyleS]

I have absolutely terrible luck, after a month I finally witnessed this bug two days in a row

[Annihilator]

Not sure what couses it, but today I had this bugg about 3-4 times... And I noticed something in the logs. All maps where bouncing props started had the same written in their error logs

L 03/12/2010 - 20:15:08: Console detour tried to unhook command "sm_dump_admcache" but it wasn't found
L 03/12/2010 - 20:15:08: Console detour tried to unhook command "sm_dump_handles" but it wasn't found
L 03/12/2010 - 20:15:08: Console detour tried to unhook command "sm" but it wasn't found

Any idea if its connected? Or how to fix it?

[rhelgeby]

These commands are not relate to the physics engine at all. We actually do know how it happens, which is stated in the first post. But not exactly when.

That ES script mentioned a few posts ago (http://addons.eventscripts.com/addons/view/bnb) do the same as we do, but on the touch event. I'm just sharing a thought now, but could this issue be related to timing? Like we change collision group too fast. Maybe we should wait for the next frame? Or maybe we're actually too late and have to change collision group as early as possible?

I've also seen people mentioning prop_physics entities in maps causing this. Maybe players in no-block mode have to touch these while dying to trigger it.

[Greyscale]

We don't know what is causing it, it is not reproducible on all servers.

I have been testing all versions of ZR on my local windows server for at least 3 years now. I have NEVER seen the bug. I have tried to reproduce it with all the scenarios that claim to do it.

One of them is "If a colliding entity is removed from the world, then the mayhem happens" Ok, so tell me what entities are being removed when anti-stick kicks in? Are players disconnecting 0.1 seconds after infection? I highly doubt it. (Not to mention I tried disconnecting while clipping through someone)

Personally, I've given up on this. I can't find any other way to unstick people (besides teleporting) and this bug is so unpredictable I can't see any possible way of working around it with the method we use.


Any of you who are getting the bug, post your server specs. Maybe it's related. I don't know what we can do about it, but if you're still interested in finding the cause just for kicks, post it here.

[raydan]

bouncing props bug happen long time,
first, i haven't run ZR, i run my own zombie plugin
but this bug sometime happened in my windows server, about 1-2 month
my plugin used some touch hook.

ZM 2.0, ZR use touch hook too, when i using ZM 2.0 (long long time ago), it also happen. All props are bouncing around the map

If someone know which part of code cause this problem, engine.dll, server.dll or vphysics.dll. may be able to use some hook or hack to fix

[Juujuu]

The code does not cause this problem

// Zombie-Related Server Variables
// ----------------------------------------------------------------------------

// Amount of time, in minutes, zombies have to infect all humans.
// Default: "5"
mp_roundtime 6

// Allows players to move items with their 'use' key.
// Default: "1"
sv_turbophysics 1

// Multiplier for how easy props are moved ['1' = No push]
// Default: "3"
phys_pushscale 3

If you have sv_turbophysics 1, then it will remain as players can move props. But, if there is a map... such as zm_natalyas_ship_v3 which forces sv_turbophysics 0, then the server will create bouncing props until the server is restarted. Sometimes you get lucky and the map doesn't change it tho.

I highly recommend removing maps which force physics to 0, the map may be fun, but chances are someone won't be online to restart it at that moment.

[rhelgeby]

I don't think this problem is related to turbophysics, because it happen both when enabled and disabled.