DescentBB

[RIP]Gallows wrote:With this AntiCheat software running I've noticed a few things occur that are undesireable.

1.Vauss doesn't seem to work as well. This has been true across the board with all the AC servers.

2. Sometimes weapons don't pick up.

3. Double mega bounce + trichord and rolling will still get you kicked.

4. People still call you a hack when you kill them.

This does not happen if you have less than 2% loss on your connect. Maybe you should fix that before posting something not really related to the A/C servers but just your own connection

Update

After many requests I finally got around to adding code to enable/disable smoothing. I also added a new feature that kicks players based on their PPS setting. This feature is able to be changed as well.

Hopefully, that'll kiss the skippys goodbye.

Very nice ! It's a bit of a double edged sword -- I know at least two 56k modem players that may have trouble getting into a server w/ their connects. I wonder if you could make it two staged, ie. <= x pps --> kick, <= y pps --> force them into a Pyro where x < y. I woud set that to x = 5 and y = 7.

Well, it is changeable by the server op, so if you wanted to set it at 5, you could.

I don't know if I could have it work based on players or change their ships. Would be interesting thought though.

I hope the PPS is related to Ping
so forth and so on 100/7, 250/5

Copied from ServerOps forum...
http://www.descentforum.net/ServerOps/v ... .php?t=243

$minpps <num>: This sets the minimum packets per second that a player must have to remain in the game. The various PPS rates are(also found in the D3 FAQ)...

28.8K modem : 5PPS
33.6K modem : 6PPS
56K modem : 7PPS
Single channel ISDN: 8PPS
Cable modem: 9PPS
T1: 12PPS

So, if you were to set minpps to 7, that would kick anyone who had the setting lower than 56K or who could not maintain 7pps in game. Now, what does it mean to say "can not maintain". Well, D3 keeps track of how many packets a player drops and if they drop enough, it will scale down their PPS. If they were at 7PPS, and they dropped enough, D3 would scale them back to 6. If you had minpps set at 7, they would be kicked.

Now, I realize yesterday servers were kicking people left and right and I sincerely apologize. I had added a double check that averaged players PPS, but observer mode, typing in <F8> mode, and even just remaining dead for a while would trip the double check. The double check is now removed and there should be no further issues with players who are actually at the rates set above.

$smoothing <on/off>: The current method of smoothing is a hybrid of D3's smoothing and what I call Positional smoothing. Each time a player sends positional data to the server, anti-cheat records it and subtracts from the previous position. It then takes the time difference and calculates a velocity from this. It then averages this velocity with the players reported velocity.
I have tried positional only smoothing and found through much testing with friends of mine, that it seemed to make gameplay smoother and provided IMO, the best playability. Unfortunately, there were some things that weren't correct like when you would bump into another player. They wouldn't move for quite some time if at all. So, I resorted back to Positional + Velocity(D3's built in) as this seems to provide the smoothness of positional without the bugs.

How to use these commands
Any server running anti-cheat can use the commands at any time either through the D3 console(on the server machine), or through a remote console(telnet).

I hope this clears up any confusion that may surround these features and provide the information necessary to use and implement in D3 server possibly. My feeling is that some kind of Post Command box could be made where one could enter additional commands that would be sent to the server after the level starts.

Bubbalou wrote: This does not happen if you have less than 2% loss on your connect. Maybe you should fix that before posting something not really related to the A/C servers but just your own connection

Considering that my loss was at 0 percent and I didn't have a problem with it on any other server, I thought I'd add my 2 cents in. Plus, on top of all of that, supersheep asked my team for feedback so we played for a couple of weeks and I submitted a small report of our findings. That's not saying the AC software sucks, its just providing the feedback that I, and many other pilots, have come up with.



AND on a side note, SuperSheeps recent updates fixed these problems so it WAS the AC software and he acknowleged that thank you very much.

Geez.

Hey SuperSheep, where can we download this software. A Few RIPs want to use it on their servers as well.


And keep up the good work. The AC software is working REAL smooth now.

[RIP]Gallows wrote:Hey SuperSheep, where can we download this software. A Few RIPs want to use it on their servers as well.


And keep up the good work. The AC software is working REAL smooth now.

See http://www.vonclan.net/Games/D3AntiCheat.htm for more information.

Update

Fusion bug fixed

When your ship and another players ship are touching and you both get with any state of charge fusion, you will both die, irregardless of your shield count in most cases.

This is most definately a bug and not a feature as some have been stating as the network traffic actually spikes during this event as well.

We have tested the fusion bug extensively and it is not as difficult as some have stated, just fire fusion as quick as you can into a crowded room, downstairs in stadium or halcyon flag rooms come to mind, and watch the death count. New players spawning in most cases don't stand a chance as they spawn in the middle of the chaos.

If you have a team game, this bug can be exploited even further by having another team member actually run against players while you both are shot. The team mate lives, but the other player/s die.

The fix actually removes the damage from the duplicate hits, and the network trafffic with it, thus making gameplay much smoother in heavy fusion games.


I am going to leave this one in because...

1) If it only happened rarely, as some have stated, then what point to leave it in? And if it happened a lot, then it impedes gameplay.
2) Network traffic is greatly reduced when the bug occurs as the duplicate damage packets are not sent.

Edited portion that stated this was fixed in 1.5.

You said this removed damage from duplicate hits. Does this mean that when fusion passes through one ship and hits another ship behind it, only the ship in front will register damage?

No, it removes duplicate hits to same player. The bug actually cause multiple damage packets to be sent from the same blob to the same player, when only one should be sent. This removes the duplicates.

Ok cool. I just misunderstood the first reading.

This removes the "fusion-bumping" exploit as well, BTW

Grendel wrote:This removes the "fusion-bumping" exploit as well, BTW

YEA!

I was wondering what the heck was going on!

The updates to AntiCheat are better seen by going to...
http://www.vonclan.net/Games/D3AntiCheat.asp

This thread should still be used for discussion of anti-cheat.

Awesome job, man now just disable the FF. ;p

Behemoth wrote:Awesome job, man now just disable the FF. ;p

that is the ServerOps discretion, talk to them!