• Please make sure you are familiar with the forum rules. You can find them here: https://forums.tripwireinteractive.com/index.php?threads/forum-rules.2334636/

Alot of NICE feats for UT99!?!

Jump to latest Follow Reply
Status
Not open for further replies.
Sort by date Sort by votes
#41
You got to be kidding right? No seriously, what kind of BS is that?

The overhead used for bullet penetration is minimal. Calculating bullet energy at impact is a ridiculously simple operation, especially if you have a good external ballistic system. Once you have the energy, you can use lookup tables or similar to determine the penetration depending on materials and projectile types. And even if you don't have a crappy (or none at all) it's easy to implement bullet penetration that is at least as good or better then the ballistic model itself without much more overhead.

Overall, the functions evaluating penetration and ricochets can be implemented in UT with negligible overhead on an engine dating back to 1999... That kind of processing was done in the UT engine, in Infiltration, for many years now.

And you are saying you cannot find a way to do the same in with out modern engine?

There's nothing neither magical nor complex about penetration. You don't even need a lot of creativity to think about ways to implement it.

From my experience with the UT engine, there is no reason not to have a good penetration system in a tactical shooter employing this engine in 2006. And "because of the ballistic model", it's probably the worse piece of BS reason I've seen in a long time. I hope this is not the official "reason"...

--edit--
Sorry for the somewhat off topic post, but when I read insanities like that, I cannot just stand there and say nothing.
 
Last edited:
Upvote 0
#42
Yes, when using hit scan weapons and the original UT code, it isn't that hard. The same code is not in UE 2.5 and we do not use hit scan weapons. Every round is caclulated all the way to the target and it takes up alot of overhead when you have things like MG 42's firing off 1200 rpm's with other high velocity guns all at once.

And watch your tone. Our insanity is what makes the game work.
 
Upvote 0
#45
geogob said:
You got to be kidding right? No seriously, what kind of BS is that?

The overhead used for bullet penetration is minimal. Calculating bullet energy at impact is a ridiculously simple operation, especially if you have a good external ballistic system. Once you have the energy, you can use lookup tables or similar to determine the penetration depending on materials and projectile types. And even if you don't have a crappy (or none at all) it's easy to implement bullet penetration that is at least as good or better then the ballistic model itself without much more overhead.

Overall, the functions evaluating penetration and ricochets can be implemented in UT with negligible overhead on an engine dating back to 1999... That kind of processing was done in the UT engine, in Infiltration, for many years now.

And you are saying you cannot find a way to do the same in with out modern engine?

There's nothing neither magical nor complex about penetration. You don't even need a lot of creativity to think about ways to implement it.

From my experience with the UT engine, there is no reason not to have a good penetration system in a tactical shooter employing this engine in 2006. And "because of the ballistic model", it's probably the worse piece of BS reason I've seen in a long time. I hope this is not the official "reason"...

--edit--
Sorry for the somewhat off topic post, but when I read insanities like that, I cannot just stand there and say nothing.
Click to expand...

Collision between game objects (not just hit scan traces) is one of the most CPU intensive operations that a game engine can do. Yes doing bullet penetration with hitscan weapons doesn't require too much overhead. But doing it with a full ballistics system does take a lot of overhead. Right now every shot fired only has 1 collision. Now you add richochets and penetration and you have 1-X(3,6,10, whatever) collision per bullet. Now multiply that times a couple of MG42s firing 1200 RPMs and 30 other players firing semi auto and full auto weapons and you have a LOT more CPU overhead. Unlike infiltration, RO not only has to run well with 32 players, but also 32 players plus any number of vehicles (8, 16, etc). That adds a LOT of server side CPU overhead as well.

So the point being, we're already at 100% server CPU utilization for a 32 player server with vehicles on a decent server. We have no headroom to add even a little additional CPU usage. Its not like we don't want penetration, and its certainly not that we don't have the skill to add it. At some point in the future if we optimize the server CPU utilization enough, we might add it. Until that point, you can certainly code up a penetration mutator and request that the servers that run with it reduce thier player counts from 32 down to 26 or something.
 
Upvote 0
#46
Your assumption that I am talking about penetration in the context using the basic UT hit scan is false.

Every round is caclulated all the way to the target
Click to expand...

This is exactly the context I had in mind when I posted my preview remark. This is also the only context in which I experimented with penetration and ricochet in the UT engine.

Spawning thousands of projectile in a period of a few seconds (few being less the 5), with penetration calculation, multiple ricochet, material detection, complexe bone-like detection on players with multiple impacts, and complex ballistics (including drop due to gravity, speed change due to drag, etc) had, during experimentation, negligable effect on the server. All calculations are done server side.

In UT.

2 years ago.
 
Upvote 0
#48
geogob said:
Your assumption that I am talking about penetration in the context using the basic UT hit scan is false.



This is exactly the context I had in mind when I posted my preview remark. This is also the only context in which I experimented with penetration and ricochet in the UT engine.

Spawning thousands of projectile in a period of a few seconds (few being less the 5), with penetration calculation, multiple ricochet, material detection, complexe bone-like detection on players with multiple impacts, and complex ballistics (including drop due to gravity, speed change due to drag, etc) had, during experimentation, negligable effect on the server. All calculations are done server side.

In UT.

2 years ago.
Click to expand...

Ok, so just add on top of that the Karma physics for vehicles (which eat a massive chunk of CPU), realistic momentum on player physics, highly advanced per poly hit detection on most objects, and any one of the hundreds of other features RO has and that "negligible effect" becomes not so negligible. Some of us have checked the effects of things in THIS engine 2 hours ago, not 2 years ago :) Your welcome to code up a mutator though if you'd like :) Thats the beauty of an moddable engine.
 
Upvote 0
#50
zeep said:
Well i for one would rather see 26 player servers WITH penetration than 32 player ones without!

Here's hoping for that mutator!
Click to expand...



As ramm said, that is the joy of a fully modable engine. Want a feature we don't agree with or choose not to implement for whatever reason? It can be made by you (and or a coder you know)!
 
Upvote 0
#51
Ok look. Nevermind.

If you want to fill your community with technical bull****, go ahead.

Trying to swim your way around the issue is not very elegant.

My "2 years ago" comment was, as I'm sure you figured that out by yourself, to point out that penetration and advanced ballistic calculation were possible with computer system considered low-end nowadays, on a very old engine that has poor optimization compared to modern engines. Overall, the impact, processing wise, of doing penetration calculation per projectile would be equal if not lower on Unreal 2.5. With such basic mathematics that use little to none special engine features, there is no reason it would be anymore CPU intensive on Unreal 2.5.

In all honesty, I think your approach in using technical information most people do not understand or are not aware of to pass your ideas around is totally unethical.

As for modding RO, the last time I checked, it was nearly 100% closed source and nearly impossible to mod. At that time, you clearly stated that you didn't want community add-ons (like there has been in Infiltration - this was the example that was stated at the time.). Of course, I'm sure the MSUC is also a great excuse to obfuscate your code. That was the first reason I removed RO from my computer system. Considering the comments I read now, the approach to modding didn't change much in the RO team, so I see no reason why I should waste my time anymore.

As a closing comment, I'd like to say that I am not the coder behind Infiltration's penetration and ballistic system, but I intensively worked on experiments and improvements of this system as well as on an experimental advanced vector-based ballistic system going as far as taking into account wind and air density.

I guess everyone has to trust your word when you use technical knowledge they do not have as an argument. Trust exactly what I totally lost when I read what you say here and elsewhere.
 
Upvote 0
#52
We have never stated we do not want community add ons. Sir I think it is you who are the one who is using mis information here and being unethical. RO is 100 % moddable. Some code was brought out of uscript format to run nativly to be more effiecient, but that is the only thing. You honestly think we don't want to add this feature we have said we wish to add and have attempted to at several times?

And if fully explaining our reasonings (or as you call it bull crap) is unethical, I don't want to see the business school you went too.
 
Upvote 0
#53
This was clearly said to me by developers, on these forums, around the time RO moved from UT2003 to UT2004. To bad the forums crashed/changed/were deleted so often.

Maybe your memory is what's failling. But I remember, oh I do!

I said RO was closed at the time I looked at this game. It may very well have changed with the latest versions. Good for you and your community. This would be, I believe, a very good move from the part of the dev team. But the comments made then were enough for me to uninstall the game and go away. Maybe I should have stayed away.

Furthermore, explaining your reasons is quite fine, but using false or flawed technical information to support your decisions is unethical. I'd have more respect if you'd simply say you are not interested in implementing the feature, or do not have the resources (time, coders, money, whatsnot) to do so.
 
Last edited:
Upvote 0
#55
You do recall the mod was under different leadership then? And even if that was true, now is much different from then. We have gone out of our way to activly encourage modding and mutators. I have gone to most of the major UT2k4 modders and asked him to take a look at RO:O as a base for their mods and mutators.

We have offered our support to them with any problems they have had, we have answered all the questions they have asked, we have even asked them if they needed features. So I am going to have to call you on this one again sir.

As to flawed arguments. If we are flawed, explain how. We are telling the truth as we know it. We will NEVER knowingly lie to our community. As I said in my other post, we have several times during the development in Ostfront attempted to add penitration. Just look at some of ramm's previous posts. He does come right out and say that he doesn't like features, or feels they were implented bad in other games.

Yes, we are a small team, yes we have limited resources and budget. And as I stated before our want list for the game is probably larger then anybody elses.
 
Upvote 0
#56
Firstly I am glad to see some long answers from TWI :)
I've been curious on how you work, and what your "collective mind" is like...
I kinda feel that people would be abit more careful bout their posting and ideas if they knew what you've been thinking behind the curtains....
For example I've been posting alot of ideas that TWI knew they would halt until UT3 no?
---------------------------------
Regarding this discussion about Infiltration and tac realism
you seem to boil down to:

1) All new feats will be stored for the UT3 engine
2) All existing feats may be tweaked or polished.
Am I correct?
-------------------------
Regarding all theese offensive and defensive arguments:
What about asking Geobob to create a penetration system that works for RO.
Let him or somebody test it, and send it to you for further evaluation?
I mean if he so clearly knows that it can be done, and how it can be done....

It seems like there are plenty of people with resourses here in the community,
but I havent seen you ask anybody "What do you mean by XYZ, and how would you do this and that"...

If I have overlooked this, I am wrong and sorry about the statement,
but if im correct, wouldnt it be a good thing to team up with the community abit more,
and **** them for the knowledge they claim to have?
----------------------------------------------
I am glad for your insaneness which made this game come true.
I wouldnt waste my time here if I didnt have faith in the game and your willingness to perfect it.

Teq
 
Last edited:
Upvote 0
#57
That would be the old Mod forums - and a long time ago, obviously. We didn't really want people to mod-the-mod, because it would all get silly.
Click to expand...

How would it be it get all silly now? Because RO is not a total conversion but a stand alone game, it would not get all silly? Or is it silly to mod a total conversion?

Looking back at Dyslexi's article about Infiltration, he didn't think the mods of the mod made it all silly... of course, you'll always have silly mods. But that doesn't make the game all silly.

now is much different from then.
Click to expand...

Again, very good for you. I'm glad for the RO community as I'm sure it is going to help improving the overall experience in the future.

Now, lets get back technical.

The worse part of penetration is really material handling. A first experimental penetration system to test would be one without material handling (because, of course, steel and wood do not have the same properties).

Having the bullet trajectory normal and speed is all you need to start with a basic penetration system. Both of these informations should be available without any calculations from a basic ballistic system. The rest is about 10 high level mathematical operations (excluding the material and bullet type lookup) which would give a little more overhead depending on the way it is implemented. Using some approximations you can simplify and optimize both the lookup and calculation very easily.
 
Upvote 0
#59
geogob said:
Now, lets get back technical.

The worse part of penetration is really material handling. A first experimental penetration system to test would be one without material handling (because, of course, steel and wood do not have the same properties).

Having the bullet trajectory normal and speed is all you need to start with a basic penetration system. Both of these informations should be available without any calculations from a basic ballistic system. The rest is about 10 high level mathematical operations (excluding the material and bullet type lookup) which would give a little more overhead depending on the way it is implemented. Using some approximations you can simplify and optimize both the lookup and calculation very easily.
Click to expand...

And we will continue to look into it and try to implement it. We are not perfect, we do not know everything. It is one of the reasons we are trying to hire another experienced coder or two. As ramm said, we are looking for ways to improve our code and reduce or overhead to give us a better chance at doing it right. We really do not have much to work with at all at the moment.
 
Upvote 0
#60
Teq said:
Regarding all theese offensive and defensive arguments:
What about asking Geobob to create a penetration system that works for RO.
Let him or somebody test it, and send it to you for further evaluation?
I mean if he so clearly knows that it can be done, and how it can be done....

It seems like there are plenty of people with resourses here in the community,
but I havent seen you ask anybody "What do you mean by XYZ, and how would you do this and that"...

If I have overlooked this, I am wrong and sorry about the statement,
but if im correct, wouldnt it be a good thing to team up with the community abit more,
and **** them for the knowledge they claim to have?
----------------------------------------------
I am glad for your insaneness which made this game come true.
I wouldnt waste my time here if I didnt have faith in the game and your willingness to perfect it.
Teq
Click to expand...

Sorry I edited too slow and u answered before finished:p
 
Upvote 0
Status
Not open for further replies.
Top Bottom
Back
Jump to latest Follow Reply
Jump to latest Follow Reply
Community News
Forums
Merch
Menu