INF style projectile penetration mutator available!

[Ron]

Ok beppo/geogob i take stuff like wooden boxes and outdoor lavatories are treated as solid pieces of wood not hollow like in rl. Is there any way you could give boxes and stuff that wont be solid pieces of wood a lower resistance to penetration sure alot of these boxes will have stuff in them but you could a value to take that into account?

 

[Omar The Insurgent]

IMHO penetration works great as it is, it is up to mappers now to take this mutator insight while making penetration specific maps.

 

[geogob]

I wouldn't base any maps on this mutator. Just make the thickness of the walls realistic.

But mappers should make sure the material surface property for all texture they use (especially the custom ones) are set properly.

Those surface properties are what is used to select which hit effect to use in RO and what kind of material the wall is made of for penetration.

 

[The_Countess]

Ok beppo/geogob i take stuff like wooden boxes and outdoor lavatories are treated as solid pieces of wood not hollow like in rl. Is there any way you could give boxes and stuff that wont be solid pieces of wood a lower resistance to penetration sure alot of these boxes will have stuff in them but you could a value to take that into account?

thats pretty eazy, but the mapper needs to do it, he can just make them hollow.
dusnt add any visable polygons and should add any significant amount to the map size, so no reason not to do it.
thats how it was done in infiltration aswell btw.

 

[Beppo]

In addition to hollow boxes you can select the proper material named HollowWood instead of Wood. But in general the mapper should model it the way it's meant to be.
So, if the box should be hollow, then he will make it hollow. If not, then it will be a solid one. Just depends on the mapper, really.

 

[Pagan]

We ran a test tonight 1 round Mut on, one round mut off, twice on odessa. I and many other people experienced a far smoother game with the mut off ie on my side, fps maxed out 80+ without mut, 30 fps with. I have to say though I love this mut I just wish it ran smoother.

 

[Pagan]

We then ran ogeldow no lag or stutter with mutator on.

 

[LordKhaine]

Well I was there too, and I was one of the people who experienced no lag or fps issues at all. Hopefully any issues get worked out because this mod adds an amazing amount to the game.

Had some really great moments today. Like when a german took cover in a barn. I emptied my SVT into it and our DP guy raked it with bullets, totally annihilating the guy inside. Or when a German went up the tower so I charged in and emptied my SVT through the wooden ceiling to clear the floor. Just makes the game far more realistic, and realistic in a fun way.

Really hope to see more servers running this mutator.

 

[ANG3L]

We ran a test tonight 1 round Mut on, one round mut off, twice on odessa. I and many other people experienced a far smoother game with the mut off ie on my side, fps maxed out 80+ without mut, 30 fps with. I have to say though I love this mut I just wish it ran smoother.

With regards to the testing of the Mutator on the server, it would appear that they is no additional workload offloaded onto the server but to the client side's instead, thus the decrease in the overall game performance on the clients side.

But this will greatly vary from client to client depending on the rig that they have, they was a drop on my system but not as noticeable in the same sense as Pagan has described on his.

What I did note at my end during the testing session was;

1) The busy-er the server got the the less likely my shots found there mark.
2) During Tank skermishes, the tanks projectiles start to behave strangely, and by this I mean;

A IS2 Tank shell penetrating a mark four up to 3-4 plus times at a range roughly around 15-30 meter's and still not killing the tank, but instead killing the man/men inside. But when the mutator was off this did not happen.

I have being drawn to both these very same conclusions on separate occasions during the short time at which our server has being testing this mutator, and the sole reason to why we are posting is allow a form of feedback to both Beppo & Geogob with regards to the mutator's behavior.

Nice job btw.

 

[Phoenix-D]

The workload being on the client's side is worrying- it opens up a potential avenue for cheating (eg MGs that can fire through the entire map without issue..)

 

[Beppo]

Guys, the reasoning and assumptions get a bit off here
Especially the last comment about cheating with MGs firing thru the whole map... I don't get what this has to do with the stuff described or with anything else... anyway...

Forgot to say 'thanks' for the feedback

Look at it this way:
Each projectile is responsible for one impact effect and one decal drawn at the impact location normally.
If the projectile is able to penetrate then you get this doubled cause you get the same decal and impact effect showing up on the backside too.
As long as you do not see this effect it will not be rendered on your system and so will not need any performance.
So, only the shots that you can see on the impact side AND the backside of the wall/object will be rendered on your machine.

So, in most situations in which you fire at a wall with no visible inside of the room behind it you will not get much of a performance loss cause the effects shouldn't be rendered at all. Sure they are spawned, but the engines relevancy check should make sure that they will not show up on your client if they aren't relevant to you.

So, the actual performance decrease for projectiles penetrating that you can only see on the first hitlocation and not where they end up or whatever number of additional walls they may penetrate is minimal.
The calcs that are done partially on the clientside too will decrease the overall performance in a very limited way.

Due to the penetrating projectiles and the mutator spawning real projectiles for short distances below the 22 meter mark too, you simply get more projectiles that show up at the same time that are relevant to your client.
So, you get more effects to show up if the stuff is in view.
In short... the more effects, the more performance is needed. This then is heavily based on your rigg and your choosen video options.

If you want to play RO with everything on ultra high, then you need a PC that is able to handle this. If you get performance issues due to some more effects showing up, then try to tune down your video options, especially the effects and physics options. They eat up a lot of performance.

Ragdolls are good looking, but you do not need ultra high physics for something that is only visual while playing online. Especially if your PC is not able to handle it fully. So, tune the physics option down first whenever possible.

 

[Pagan]

Later on we ran baksan and a few other maps with a half full server and everything ran great, we will run more tests with different options to give more feedback.

 

[Phoenix-D]

Guys, the reasoning and assumptions get a bit off here
Especially the last comment about cheating with MGs firing thru the whole map... I don't get what this has to do with the stuff described or with anything else... anyway...

Simple- if the lag is client side, that sounds like the calculations are being done client side. If the calculations are done client side is possible to alter them and cheat- hence the "Mg firing through the map" comment.

 

[zeep]

Simple- if the lag is client side, that sounds like the calculations are being done client side. If the calculations are done client side is possible to alter them and cheat- hence the "Mg firing through the map" comment.

Whats done clientside are the graphics. Servers do all the penetration calculations so there can't be any "Mg firing through the map" cheating. Unless such a bug exists in the map itself, in which case it would be possible with- or without the penetration mutator.

 

[ANG3L]

Look at it this way:
Each projectile is responsible for one impact effect and one decal drawn at the impact location normally.
If the projectile is able to penetrate then you get this doubled cause you get the same decal and impact effect showing up on the backside too.
As long as you do not see this effect it will not be rendered on your system and so will not need any performance.
So, only the shots that you can see on the impact side AND the backside of the wall/object will be rendered on your machine.

So, in most situations in which you fire at a wall with no visible inside of the room behind it you will not get much of a performance loss cause the effects shouldn't be rendered at all. Sure they are spawned, but the engines relevancy check should make sure that they will not show up on your client if they aren't relevant to you.

So, the actual performance decrease for projectiles penetrating that you can only see on the first hitlocation and not where they end up or whatever number of additional walls they may penetrate is minimal.

The calcs that are done partially on the clientside too will decrease the overall performance in a very limited way.

Due to the penetrating projectiles and the mutator spawning real projectiles for short distances below the 22 meter mark too, you simply get more projectiles that show up at the same time that are relevant to your client.

So, you get more effects to show up if the stuff is in view.

In short... the more effects, the more performance is needed. This then is heavily based on your rigg and your choosen video options.

Maybe then you could drop the radius of the viewable field from a spectator 22 meters to relativity closer 10-5 meters? to see how this will have an affect on the performance outcome.

My only concern is for the little guys as it where with the not so grunty systems, they is already a great deal of game code going through there CPU's and this can and will have an impact on there CPU's running the other codes in the game.

Just would like to see this great little feature not go down the road of;
"we'll if your system doesn't have what it takes then tough! route.

 

[Pagan]

I dunno why but Odessa is a ***** with this on. My settings are all at normal but when a MG on the other side of the map lights up I stutter like Claudius.

 

[geogob]

While it shutters on Odessa, check your logs for anything out of it's place. If you have two monitors, you can start RO with the -log parameter to have the log window opened on the second monitor (that's always how i test things, works great for instant feedback). Also check the server logs from that time to spot out anything usefull.

Maybe the shutter is associated with accessed nones or something like that that might point to another problem due to the fact that this is implemented as a mutator and not directly in the game.

--

As for cheating, I feel you have little knowledge on the way the engine handles such things like replication, damage processing and effect rendering. Rest assured, there is no way to exploit this. Do not worry about this. If anything is exploitable now, it was before as well. But again, rest assured... no such risk.

 

[geogob]

So no news from TWI about those native functions? It's thankgiving weekend over here in Canada. Would have been nice to have some more input about those functions...

I guess no one wants to do some overtime and copy-paste the C++ code for a function or two? (or maybe just the pseudo code so we understand what it all does in the inside and how to get around it)

 

[Dyvim Tvar]

{1.SS} Server is now running the penetration mutator for testing!

 

[geogob]

Good to read
Thx for the support.