Project Orbiter Battle Simulation Project Needs Developers! - Page 9 - Orbiter-Forum

T.Neo · 10-04-2010, 04:31 PM

Quote:

In general I'd agree with you, because converting meshes would save a lot of time and would ensure we actually had *any* meshes, other then missiles and bombs...

Yeah... I don't think I will show the results of my first attempt to mesh a tank.

Quote:

We're putting infantry in cos it's not costing us much work at this stage.

Apart from ridiculously difficult meshes of humans, unless we just want to have UMMUs sliding around shooting eachother.

Wishbone · 10-04-2010, 04:37 PM

Would you care to post any screenshots? And no, videos don't count since I cannot access them.

RisingFury · 10-04-2010, 04:45 PM

Quote:

Originally Posted by Wishbone

Would you care to post any screenshots? And no, videos don't count since I cannot access them.

Screenshots are in the first post.

Wishbone · 10-04-2010, 06:20 PM

Guess my reading skills need a bit of honing... Thanks, I can see the light at the end of the wind tunnel now. Re: infantry combat - considering machine guns (or metal storm) will be frequently used, do you plan to track each bullet or resolve hits statistically? Orbiter doesn't like very large scenarios yet.

RisingFury · 10-04-2010, 09:21 PM

Quote:

Originally Posted by T.Neo

we are going to simulate bullets as vessels.

What in hell are you talking about?! There's no way we can afford that!

Before the ground collision was rewritten and activated I started flying around with a combat DG, dropping bombs everywhere. Even with labels turned off and me zooming so far out you couldn't even see the Sun anymore, the framerate was still around 20 FPS. That was with 500 bombs laying on the ground.

I'm running an AMD 3000+ CPU and given that it's one of the weaker computers around here by now, might as well use it as a testbed...

Maybe we'll fire a few shells out every second to let people know where the bullets are heading...

T.Neo · 10-04-2010, 10:18 PM

Quote:

What in hell are you talking about?! There's no way we can afford that!

Well, that's what I heard, sorry if my statement was inaccurate due to further testing.

**Hielor** · 10-05-2010, 12:07 AM

Quote:

Originally Posted by RisingFury

What in hell are you talking about?! There's no way we can afford that!

Before the ground collision was rewritten and activated I started flying around with a combat DG, dropping bombs everywhere. Even with labels turned off and me zooming so far out you couldn't even see the Sun anymore, the framerate was still around 20 FPS. That was with 500 bombs laying on the ground.

I'm running an AMD 3000+ CPU and given that it's one of the weaker computers around here by now, might as well use it as a testbed...

Maybe we'll fire a few shells out every second to let people know where the bullets are heading...

Ah, but there's no reason to have 500 bombs laying on the ground--once the bomb gets to the ground, you can have it delete itself (after applying whatever effect it would've had).

And yes, an AMD 3000+ is on the low end at this point in time, but better to make it work well on low-end hardware than have it work well on high-end hardware and be unusable on low-end hardware.

So since you won't be simulating individual bullets, you're going to be using some kind of statistical determination of whether a hit is scored? That question has been asked a couple times in the thread, but I haven't seen an answer (sorry if I missed it, the thread sometimes goes through a dozen or more posts between when I get to reading it)

RisingFury · 10-11-2010, 05:22 PM

Quote:

Originally Posted by Hielor

Ah, but there's no reason to have 500 bombs laying on the ground--once the bomb gets to the ground, you can have it delete itself (after applying whatever effect it would've had).

And yes, an AMD 3000+ is on the low end at this point in time, but better to make it work well on low-end hardware than have it work well on high-end hardware and be unusable on low-end hardware.

So since you won't be simulating individual bullets, you're going to be using some kind of statistical determination of whether a hit is scored? That question has been asked a couple times in the thread, but I haven't seen an answer (sorry if I missed it, the thread sometimes goes through a dozen or more posts between when I get to reading it)

Sorry for the late reply, didn't notice your post...

Well, of course missiles and bombs will get deleted once they're done blowing things up. I was just pointing out that we can't afford such high numbers of vessels in game and if you're talking about a multi-barrel gun that fires many hundred rounds per minute, it could get really bad quickly...

We can't just go with statistics based and we can't use Orbiter's vessel class, because it's too big for our requirement... So at this point I see three choices...

- Create a 'lite' class that gets fired like a bullet, but isn't derived from VESSEL. Propagate that. That might be the most accurate, but most resource hungry method.

- Make a system were any fast firing gun - automatic weapons, Gatling guns,... would be handled like a laser, but affected by gravity. Basically check the flight path instantly for vessels, while keeping the ability to fire VESSEL derived shells, for tanks and artillery...

- Make a system that is a bit more advanced then the last described - split the path into segments and predict if any vessel will be in the flight path when the bullet hits it. This one might be a bit more resource unfriendly, but I imagine better then the first one. It wouldn't be as accurate, but better then the second option. We've yet to discuss and debate our options though.

In either case, I'd limit the range of bullets (not shells) to a few kilometer range and have them delete themselves when crossing it...

Any suggestions (from qualified) people are welcome

RisingFury · 10-11-2010, 08:28 PM

Quote:

Originally Posted by Coolhand

sorta like a CCIP mode in aircraft fire control systems, you'll also need to account for velocities and in an atmosphere perhaps wind, certainly dynamic pressures, no?

but still you'll not want to work out the impact point for every single round... or you don't need to... I'd have figured on simply rendering tracer, say 1 in 10 rounds? so if you're firing 500rpm then you're only calculating a maximum a 50 impacts per minute (and only drawing 1/10th of total rounds), but account for 10 impacts or a damage multiplier for each impact. perhaps give targets a wider hit-box to account for shotgun effects with a reduced number of calculated projectiles.

Quote:

Originally Posted by Hielor

I thought of another potential optimization here: ignore the possibility for collateral damage and have bullets only be capable of hitting whatever their original target was. If they pass their target without scoring a hit, they can be safely deleted (since you know they're never going to hit it).

This also saves a whole lot of math with collision detections against ships other than the primary target, at the cost of some realism--but if you're looking for performance gains, it's probably realism that you can afford to sacrifice.

Thanks! Good ideas!

Wishbone · 10-11-2010, 11:01 PM

Has anybody mentioned infantry combat here (hint - it wasn't me

)? With machine guns and he-frag arty fires? I was kinda responding to the explicitly expressed desire to have bullet-level, hi-fidelity modelling. If there's only _space_ combat, how would you model attacks on bases?

Coolhand · 10-11-2010, 11:14 PM

yeah it would be a pain in the ass if you were strafing a column of tanks and had to switch targets constantly... and how about firing with no lock at all?

perhaps if you're talking a ground war, or maybe so that system is flexible enough to support even starwars like scenarios the problem is not so much to do with stray shells with a finite lifespan (which could perhaps be tunable by the player depending on how fast their machine is) as the sheer amount of other stuff that could be going on.

Falcon 4 / Freefalcon, if anyones played that (if not get it, its freeware and its a great sim) uses a 'bubble' system, so groups of units are treated as 1 entity with attrition of individual units calculated statistically. until the player comes within a certain range only then will the machine have to think about individual units. so they can have a large war persisting over the entire country (and its a dynamic, not scripted campaign) even on the 1998 hardware it was originally designed for.

**Hielor** · 10-11-2010, 11:15 PM

Quote:

Originally Posted by Wishbone

Has anybody mentioned infantry combat here (hint - it wasn't me

)? With machine guns and he-frag arty fires? I was kinda responding to the explicitly expressed desire to have bullet-level, hi-fidelity modelling. If there's only _space_ combat, how would you model attacks on bases?

I'm pretty sure that standard battle tactics for infantry haven't involved "massed formations" since, oh I don't know, World War II?

I'm also still unsure why we need infantry combat in a space simulator. "How would you model attacks on bases?" Simple enough--simulate the infantry combat to a level good enough for a flight sim. Plenty of flight sims have simulated tank and infantry combat (for AI) without completely recreating them.

Wishbone · 11-01-2010, 09:15 PM

Orbiter doesn't know what fog of war is... sad but true... everybody's sitawareness is always perfect. Re: both ways - the side with the more numerous fleet wins, Lanchester equations work in space as well as they do on Earth.

**Hielor** · 11-01-2010, 09:50 PM

Quote:

Originally Posted by Wishbone

Re: both ways - the side with the more numerous fleet wins, Lanchester equations work in space as well as they do on Earth.

Except that "the side with the more numerous fleet wins" doesn't seem to hold true very often on Earth...

Polaris · 02-19-2011, 05:22 PM

Is OBSP ready for further testing- or is it finished?

10-04-2010, 09:21 PM	#125
RisingFury OBSP developer	Quote: Originally Posted by T.Neo we are going to simulate bullets as vessels. What in hell are you talking about?! There's no way we can afford that! Before the ground collision was rewritten and activated I started flying around with a combat DG, dropping bombs everywhere. Even with labels turned off and me zooming so far out you couldn't even see the Sun anymore, the framerate was still around 20 FPS. That was with 500 bombs laying on the ground. I'm running an AMD 3000+ CPU and given that it's one of the weaker computers around here by now, might as well use it as a testbed... Maybe we'll fire a few shells out every second to let people know where the bullets are heading... Last edited by RisingFury; 10-04-2010 at 09:27 PM.

10-11-2010, 11:15 PM	#132
Hielor Defender of Truth	Quote: Originally Posted by Wishbone Has anybody mentioned infantry combat here (hint - it wasn't me )? With machine guns and he-frag arty fires? I was kinda responding to the explicitly expressed desire to have bullet-level, hi-fidelity modelling. If there's only _space_ combat, how would you model attacks on bases? I'm pretty sure that standard battle tactics for infantry haven't involved "massed formations" since, oh I don't know, World War II? I'm also still unsure why we need infantry combat in a space simulator. "How would you model attacks on bases?" Simple enough--simulate the infantry combat to a level good enough for a flight sim. Plenty of flight sims have simulated tank and infantry combat (for AI) without completely recreating them. Last edited by Hielor; 10-11-2010 at 11:20 PM.

Quick Links	Need Help?
About Us Register to Participate Orbiter Website Orbiter Download Projects Recommended Addons Orbit Hangar Mods Photo Gallery Social Groups External Links	Did you forget your password? Rules & Guidelines Frequently Asked Questions Orbiter F.A.Q. Orbiter Tutorials Orbiter IRC Chat Search Forum Site Map Contact Us Mark Forums Read

10-04-2010, 04:31 PM	#121
T.Neo SA 2010 Soccermaniac	Quote: In general I'd agree with you, because converting meshes would save a lot of time and would ensure we actually had any meshes, other then missiles and bombs... Yeah... I don't think I will show the results of my first attempt to mesh a tank. Quote: We're putting infantry in cos it's not costing us much work at this stage. Apart from ridiculously difficult meshes of humans, unless we just want to have UMMUs sliding around shooting eachother.

10-04-2010, 04:37 PM	#122
Wishbone Clueless developer	Would you care to post any screenshots? And no, videos don't count since I cannot access them.

10-04-2010, 04:45 PM	#123
RisingFury OBSP developer	Quote: Originally Posted by Wishbone Would you care to post any screenshots? And no, videos don't count since I cannot access them. Screenshots are in the first post.

10-04-2010, 06:20 PM	#124
Wishbone Clueless developer	Guess my reading skills need a bit of honing... Thanks, I can see the light at the end of the wind tunnel now. Re: infantry combat - considering machine guns (or metal storm) will be frequently used, do you plan to track each bullet or resolve hits statistically? Orbiter doesn't like very large scenarios yet.

10-04-2010, 10:18 PM	#126
T.Neo SA 2010 Soccermaniac	Quote: What in hell are you talking about?! There's no way we can afford that! Well, that's what I heard, sorry if my statement was inaccurate due to further testing.

10-05-2010, 12:07 AM	#127
Hielor Defender of Truth	Quote: Originally Posted by RisingFury What in hell are you talking about?! There's no way we can afford that! Before the ground collision was rewritten and activated I started flying around with a combat DG, dropping bombs everywhere. Even with labels turned off and me zooming so far out you couldn't even see the Sun anymore, the framerate was still around 20 FPS. That was with 500 bombs laying on the ground. I'm running an AMD 3000+ CPU and given that it's one of the weaker computers around here by now, might as well use it as a testbed... Maybe we'll fire a few shells out every second to let people know where the bullets are heading... Ah, but there's no reason to have 500 bombs laying on the ground--once the bomb gets to the ground, you can have it delete itself (after applying whatever effect it would've had). And yes, an AMD 3000+ is on the low end at this point in time, but better to make it work well on low-end hardware than have it work well on high-end hardware and be unusable on low-end hardware. So since you won't be simulating individual bullets, you're going to be using some kind of statistical determination of whether a hit is scored? That question has been asked a couple times in the thread, but I haven't seen an answer (sorry if I missed it, the thread sometimes goes through a dozen or more posts between when I get to reading it)

10-11-2010, 05:22 PM	#128
RisingFury OBSP developer	Quote: Originally Posted by Hielor Ah, but there's no reason to have 500 bombs laying on the ground--once the bomb gets to the ground, you can have it delete itself (after applying whatever effect it would've had). And yes, an AMD 3000+ is on the low end at this point in time, but better to make it work well on low-end hardware than have it work well on high-end hardware and be unusable on low-end hardware. So since you won't be simulating individual bullets, you're going to be using some kind of statistical determination of whether a hit is scored? That question has been asked a couple times in the thread, but I haven't seen an answer (sorry if I missed it, the thread sometimes goes through a dozen or more posts between when I get to reading it) Sorry for the late reply, didn't notice your post... Well, of course missiles and bombs will get deleted once they're done blowing things up. I was just pointing out that we can't afford such high numbers of vessels in game and if you're talking about a multi-barrel gun that fires many hundred rounds per minute, it could get really bad quickly... We can't just go with statistics based and we can't use Orbiter's vessel class, because it's too big for our requirement... So at this point I see three choices... - Create a 'lite' class that gets fired like a bullet, but isn't derived from VESSEL. Propagate that. That might be the most accurate, but most resource hungry method. - Make a system were any fast firing gun - automatic weapons, Gatling guns,... would be handled like a laser, but affected by gravity. Basically check the flight path instantly for vessels, while keeping the ability to fire VESSEL derived shells, for tanks and artillery... - Make a system that is a bit more advanced then the last described - split the path into segments and predict if any vessel will be in the flight path when the bullet hits it. This one might be a bit more resource unfriendly, but I imagine better then the first one. It wouldn't be as accurate, but better then the second option. We've yet to discuss and debate our options though. In either case, I'd limit the range of bullets (not shells) to a few kilometer range and have them delete themselves when crossing it... Any suggestions (from qualified) people are welcome

10-11-2010, 08:28 PM	#129
RisingFury OBSP developer	Quote: Originally Posted by Coolhand sorta like a CCIP mode in aircraft fire control systems, you'll also need to account for velocities and in an atmosphere perhaps wind, certainly dynamic pressures, no? but still you'll not want to work out the impact point for every single round... or you don't need to... I'd have figured on simply rendering tracer, say 1 in 10 rounds? so if you're firing 500rpm then you're only calculating a maximum a 50 impacts per minute (and only drawing 1/10th of total rounds), but account for 10 impacts or a damage multiplier for each impact. perhaps give targets a wider hit-box to account for shotgun effects with a reduced number of calculated projectiles. Quote: Originally Posted by Hielor I thought of another potential optimization here: ignore the possibility for collateral damage and have bullets only be capable of hitting whatever their original target was. If they pass their target without scoring a hit, they can be safely deleted (since you know they're never going to hit it). This also saves a whole lot of math with collision detections against ships other than the primary target, at the cost of some realism--but if you're looking for performance gains, it's probably realism that you can afford to sacrifice. Thanks! Good ideas!

10-11-2010, 11:01 PM	#130
Wishbone Clueless developer	Has anybody mentioned infantry combat here (hint - it wasn't me )? With machine guns and he-frag arty fires? I was kinda responding to the explicitly expressed desire to have bullet-level, hi-fidelity modelling. If there's only _space_ combat, how would you model attacks on bases?

10-11-2010, 11:14 PM	#131
Coolhand Addon Developer	yeah it would be a pain in the ass if you were strafing a column of tanks and had to switch targets constantly... and how about firing with no lock at all? perhaps if you're talking a ground war, or maybe so that system is flexible enough to support even starwars like scenarios the problem is not so much to do with stray shells with a finite lifespan (which could perhaps be tunable by the player depending on how fast their machine is) as the sheer amount of other stuff that could be going on. Falcon 4 / Freefalcon, if anyones played that (if not get it, its freeware and its a great sim) uses a 'bubble' system, so groups of units are treated as 1 entity with attrition of individual units calculated statistically. until the player comes within a certain range only then will the machine have to think about individual units. so they can have a large war persisting over the entire country (and its a dynamic, not scripted campaign) even on the 1998 hardware it was originally designed for.

11-01-2010, 09:15 PM	#133
Wishbone Clueless developer	Orbiter doesn't know what fog of war is... sad but true... everybody's sitawareness is always perfect. Re: both ways - the side with the more numerous fleet wins, Lanchester equations work in space as well as they do on Earth.

11-01-2010, 09:50 PM	#134
Hielor Defender of Truth	Quote: Originally Posted by Wishbone Re: both ways - the side with the more numerous fleet wins, Lanchester equations work in space as well as they do on Earth. Except that "the side with the more numerous fleet wins" doesn't seem to hold true very often on Earth...

02-19-2011, 05:22 PM	#135
Polaris Orbinaut	Is OBSP ready for further testing- or is it finished?