Who would purchase a Steel beasts DCS module

[Flagrum]

I just had a crazy idea.

 

SB vehicles are modelled in more detail than DCS vehicles (especially damage model).

DCS aircraft are modeelled in more detail than SB aircraft (uhm, yes, there ARE aircraft in SB! ;o)

SB maps are more detailled than DCS maps. Not necessarily in regards to geometry (Wags spoke about 1m grids for elevation data for parts of NTTR?), but in regards to properties of the surface (although SB has there potential for imporvements as well - big time, imho).

Therefore(?) SB maps tend to be much smaller than DCS maps (civil version of SB: max 50x50 km? vs. DCS: 300-400 nm).

Weapons are modelled in quite some detail in both - but with focus on different weapons.

 

Conclusion: SB and DCS won't fit together, they are too different and it would be a huuuuge task to put both worlds together, technically - and probably otherwise as well.

 

Really?

 

What about this:

"SB 'instances' in a DCS world"? A probably multiplayer-only approach: we have a huge map for aircraft to fly over, but small areas are SB territory. A DCS server takes care about everthing flying and an SB server takes care of each SB territory and everything happening in these territories is mirrored to the other side, to the respective other server.

 

Example:

DCS Server: I fly a A-10C over a DCS map. On the ground I see a tank. I drop a Mk-82 next to him, disabling it, but not killing it. I fly home.

SB server: A enemy plane appears on the edge of the SB map, seems to be a A-10. It drops a bomb, but misses me - but still archievs a mobility "kill". Nevertheless, I keep firing at the other tanks around me.

 

The units and their movement is synchronised between the DCS server and the respective SB instance. The weapon effects are calculated on the side where it happens (in this case the SB server).

 

Basically, from a DCS perspective, most of the technology is already there. Such a SB instance could be treated similar to a MP CA user - but it act's like a proxy where several actual users are hidden behind.

 

How much work would be necessary on the side of SB ... I have no idea, though... They would probably need to implement a way that allows units to be controlled not by a user and not by the AI, but by ... something else.

 

Another problem, but solvble, I think, would be the terrain data. But SB already has some capable map editing tools, so "external" maps should be includable relatively easily. Maybe a way to deal with differences on the maps is necessary (i.e. different elevation data resolutions).

[Marko321]

I just had a crazy idea.

 

SB vehicles are modelled in more detail than DCS vehicles (especially damage model).

DCS aircraft are modeelled in more detail than SB aircraft (uhm, yes, there ARE aircraft in SB! ;o)

SB maps are more detailled than DCS maps. Not necessarily in regards to geometry (Wags spoke about 1m grids for elevation data for parts of NTTR?), but in regards to properties of the surface (although SB has there potential for imporvements as well - big time, imho).

Therefore(?) SB maps tend to be much smaller than DCS maps (civil version of SB: max 50x50 km? vs. DCS: 300-400 nm).

Weapons are modelled in quite some detail in both - but with focus on different weapons.

 

Conclusion: SB and DCS won't fit together, they are too different and it would be a huuuuge task to put both worlds together, technically - and probably otherwise as well.

 

Really?

 

What about this:

"SB 'instances' in a DCS world"? A probably multiplayer-only approach: we have a huge map for aircraft to fly over, but small areas are SB territory. A DCS server takes care about everthing flying and an SB server takes care of each SB territory and everything happening in these territories is mirrored to the other side, to the respective other server.

 

Example:

DCS Server: I fly a A-10C over a DCS map. On the ground I see a tank. I drop a Mk-82 next to him, disabling it, but not killing it. I fly home.

SB server: A enemy plane appears on the edge of the SB map, seems to be a A-10. It drops a bomb, but misses me - but still archievs a mobility "kill". Nevertheless, I keep firing at the other tanks around me.

 

The units and their movement is synchronised between the DCS server and the respective SB instance. The weapon effects are calculated on the side where it happens (in this case the SB server).

 

Basically, from a DCS perspective, most of the technology is already there. Such a SB instance could be treated similar to a MP CA user - but it act's like a proxy where several actual users are hidden behind.

 

How much work would be necessary on the side of SB ... I have no idea, though... They would probably need to implement a way that allows units to be controlled not by a user and not by the AI, but by ... something else.

 

Another problem, but solvble, I think, would be the terrain data. But SB already has some capable map editing tools, so "external" maps should be includable relatively easily. Maybe a way to deal with differences on the maps is necessary (i.e. different elevation data resolutions).

 

 

 

The Maps used by esim for there military contracts are substantially bigger Then the PE version.

[NineLine]

I just had a crazy idea.

 

I think it would be much the same as any developer from anywhere else, they would pretty much have to start from the ground up again, maybe very base things like 3D models (which would still need work) could be used, but for the most part, all they would be able to bring is their knowledge, everything else would need to be re-done...

[Flagrum]

The Maps used by esim for there military contracts are substantially bigger Then the PE version.

I know, and lager SB instances would be even cooler, but for us civil customers it probably doesn't really matter.

[Marko321]

I think it would be much the same as any developer from anywhere else, they would pretty much have to start from the ground up again, maybe very base things like 3D models (which would still need work) could be used, but for the most part, all they would be able to bring is their knowledge, everything else would need to be re-done...

 

I would agree with that.

But lets not devalue the there Data base of Armour strengths ballistics etc.

They have spent as many years studying Armour as DCS has on military

Aircraft

[Flagrum]

I think it would be much the same as any developer from anywhere else, they would pretty much have to start from the ground up again, maybe very base things like 3D models (which would still need work) could be used, but for the most part, all they would be able to bring is their knowledge, everything else would need to be re-done...

No, that is, I believe/hope, the best part of my idea: each side keeps it's assets and logic. They just look a bit differently. But the behaviour is controlled by the server that knows best how to do that. The SB instance server contolls AI units in a SB way and deals with SB user units in a SB way. The results are then mirrored into the DCS world.

 

A DCS aircraft attacks a SB ground unit - SB decides about the outcome, the weapon effect and the results for the target. This is then sent back to DCS and the pilot sees a tank explode ... or not.

 

And vice versa: I use my MG to defend my SB tank when an DCS aircraft appears on my SB server. SB tells DCS the bullet trajectory and DCS decides if and what part of the a/c was hit and what damage was dealt.

 

edit:

What both sides would need to implement is an interface where both servers can exchange information about unit positions/movement, weapon position/trajectory and a way to exchange the results. And maybe some "small stuff", like radio comms across server boundaries ...

Edited November 5, 2014 by Flagrum

[NineLine]

I would agree with that.

But lets not devalue the there Data base of Armour strengths ballistics etc.

They have spent as many years studying Armour as DCS has on military

Aircraft

 

No one was devaluing any of their knowledge, but its the ability to make all the knowledge work in DCS that becomes the hitch, they have no experience with DCS right now, so it would be like starting over...

[NineLine]

No, that is, I believe/hope, the best part of my idea: each side keeps it's assets and logic. They just look a bit differently. But the behaviour is controlled by the server that knows best how to do that. The SB instance server contolls AI units in a SB way and deals with SB user units in a SB way. The results are then mirrored into the DCS world.

 

A DCS aircraft attacks a SB ground unit - SB decides about the outcome, the weapon effect and the results for the target. This is then sent back to DCS and the pilot sees a tank explode ... or not.

 

And vice versa: I use my MG to defend my SB tank when an DCS aircraft appears on my SB server. SB tells DCS the bullet trajectory and DCS decides if and what part of the a/c was hit and what damage was dealt.

 

 

Sorry, I see what you mean now... and you are right, its crazy ;)

[Marko321]

No, that is, I believe/hope, the best part of my idea: each side keeps it's assets and logic. They just look a bit differently. But the behaviour is controlled by the server that knows best how to do that. The SB instance server contolls AI units in a SB way and deals with SB user units in a SB way. The results are then mirrored into the DCS world.

 

A DCS aircraft attacks a SB ground unit - SB decides about the outcome, the weapon effect and the results for the target. This is then sent back to DCS and the pilot sees a tank explode ... or not.

 

And vice versa: I use my MG to defend my SB tank when an DCS aircraft appears on my SB server. SB tells DCS the bullet trajectory and DCS decides if and what part of the a/c was hit and what damage was dealt.

 

edit:

What both sides would need to implement is an interface where both servers can exchange information about unit positions/movement, weapon position/trajectory and a way to exchange the results. And maybe some "small stuff", like radio comms across server boundaries ...

 

Well the Australian Armed forces have managed to integrate SB with VBS

How they manage to do this is beyond my pay scale.

 

Edited November 5, 2014 by Marko321

[cichlidfan]

If you can put in vehicles at the same level as A-10C, with realistic infantry, your sales will SKYROCKET!

 

Only if you can also write an AI that can handle all of that very much better than what we currently have.

[Marko321]

Only if you can also write an AI that can handle all of that very much better than what we currently have.

 

Not sure about the infantry.

That mite be a step to far having arma 3 type infantry would be great Though.

 

A post on another forum about this topic in regards to SB

Got a response from a guy, Who works in software development he believes its Technically possible.

Edited November 5, 2014 by Marko321

[Fri13]

I have no doubt that ED can implement Armor to a high degree of realism...judging by the A10C...my worry would be that they can do it with a terrain to match. Cause if you can't do the terrain right ...you haven't got Sh@#. Having said that a "micro terrain" is concerning as well. Imagine the potential frame rate hit that might be.

Terrain doesn't need to be a complex, like include a fluid mechanics for rain and mud, or even rolling stones. Even tracks/wheels leaving marks can be made on client end if wanted with shaders.

 

There are great examples how large a area terrain can be made so it has mesh to support fine details (like 0.5x0.5m or 1x1m) and such features that even MBT or engineer tank can dig its cover on ground.

 

The question is, what EDGE can do about that?

Can EDGE support such features to current map, or does map need to be redone or can it be just converted?

If i remember correctly, we are getting two times smaller mesh for terrain, that is the current 100x100m area comes down to 50x50m size.

 

There are so many small things helicopter pilots would love to see, like bending trees if your helicopter fuselag touch them, or trees waving by turbulence. But same manner trees bending and falling when MBT hits them.

 

The videos we have seen about MBT physics like the tensions and hull impact from firing gun, same thing would be great for trees.

 

I just bought a old T-72/T-55/T-34 "Sim" from steam and it had fancier terrain physics than I thought. You actually see .5m mesh and it affects to vehicle movement very well. It was very believable driving experience from so simple modeling.

 

First time when I saw a MBT simulation, it was Microprose M1 Tank Platoon. That time beat flight simulator I had seen (commercial) was Microprose FighterBomber.

Then came later M1 Tank Platoon II and Gunship! and those two were designed to work together in multiplayer (never was patched). Since then I have wanted to see such combo with fancier terrain.

Until 2001 Operation Flashpoint was released.

 

Even today I am amazed about OFP and ARMA II terrain, while being simple, it does presents nice landscapes with rocks etc.

If we ever get ARMA II level terrain, it will be huge improvement.

 

And lots of problems can be hided with small details, like rocks, individual trees and bushes etc obstructing line of sight.

Now in DCS it is forest, building or terrain that blocks the view. Few treelines that are very nice but not enough.

 

And many such small details could be rendered as well only on client side, like small rocks and high grass as the normal grass is.

[tobaschi]

i love steal beasts pro, it's one of the simulations that i play quit regular and for sure i would buy a DCS ground modul if it would be crafted by esim games, because it would be a gurantor for high quality

[Zakatak]

I didn't want one before, but with the addition of EDGE, the game should look decent enough from the ground to make a proper tank simulation. Aside from Steel Beasts and Red Orchestar 2, you can't really get any good tank experiences as of now. Preferably a 3 crew tank to give the player as much agency as possible, like the T-90 or Leclerc.

[theGozr]

Yes cockpits we need and it will bring lots of people :)

[winz]

Well the Australian Armed forces have managed to integrate SB with VBS

How they manage to do this is beyond my pay scale.

 

How did they managed it? Easy, they sad they wanted it, the company(ies) analyzed the request, made an estimate how much it would cost, made an offer to which the australian armed forced agreed and payed the whole development. ;)

 

What we don't see in that short video is how good is the integration in general. The thing to keep in mind is that big customer (like militaries), who pay for the whole development, are usually focused on some aspect of the simulation and don't care much (or don't intend to fund) aspects that are of no value to them (like AI, damage modeling..etc). So while the integration might 100% work in some aspects, like syncing VBS + SB unit position, it might be totally broken and non-functional in other aspects.

 

One immediate issue that comes to mind when you try to merge flight sim with a infantry sim is the issue of different scenery fidelity. The problem Arma has(d) with grass just becames 100x worse. You just cannot render the whole visible area of a flight sim in the fidelity that is required at the infantry sim level. So you end up with situations where the infantry player is hiding behind scenery objects that are not rendered for the pilot player, thus making him an easy target. And this is a huge usability issue that is very difficult to solve.

[Marko321]

How did they managed it? Easy, they sad they wanted it, the company(ies) analyzed the request, made an estimate how much it would cost, made an offer to which the australian armed forced agreed and payed the whole development. ;)

 

What we don't see in that short video is how good is the integration in general. The thing to keep in mind is that big customer (like militaries), who pay for the whole development, are usually focused on some aspect of the simulation and don't care much (or don't intend to fund) aspects that are of no value to them (like AI, damage modeling..etc). So while the integration might 100% work in some aspects, like syncing VBS + SB unit position, it might be totally broken and non-functional in other aspects.

 

One immediate issue that comes to mind when you try to merge flight sim with a infantry sim is the issue of different scenery fidelity. The problem Arma has(d) with grass just becames 100x worse. You just cannot render the whole visible area of a flight sim in the fidelity that is required at the infantry sim level. So you end up with situations where the infantry player is hiding behind scenery objects that are not rendered for the pilot player, thus making him an easy target. And this is a huge usability issue that is very difficult to solve.

 

Straight From the horses mouth.

This post is quoted from the senior Australian officer who oversees SB

 

Yes this has been problematic for us (unsure if its purely technical or lack of will on a side) with VBS and another product near and dear to our hearts via the LVC game route.

 

Once you get the object library sorted, you then move onto issues about collated terrain, etc.

 

And that's with two ground force products.

[NineLine]

Guys we need to keep the conversation mostly on DCS and Combined Arms, SB and other sims have their own forums, its not likely we will see any joint projects between ED and eSims, least anything that is public knowledge, so its not really a constructive discussion at this point.

[dumgrunt]

terrain detail is not the problem. the current map has OFP LOD essentially, which considering the the scale and the purpose (air sim) is pretty bloody good.

 

the main issue is accurately simulating ground warfare, in particular the relationship between the various arms, little lone tactics, protection, weapons effects, then in addition the real world detailed stuff like active and passive protection measures, concealment and cover. I could go on an on...

ARMA is leaps and bounds ahead of DCS in this regard and it is (in term of DCS air vs AMRA ground) arcade compared to DCS.

 

Its no bloody good having a DCS A-10C level abrams set in a "street fighter '89" standard battlefield. you would be pissing into the wind.

[TwilightZone]

If such a module were ever to be made available

I was wondering would it be received well by DCS players

Combined Arms is good at what it was designed for.

But as we all know its not a high fidelity armoured simulation.

.....yeah.....i'd be up for a tank sim done by ED:thumbup:

[FlankerKiller]

I agree the terrain detail is good enough, if we can get collidable trees. All of the AI upgrades you mentioned would massively improve the single player experience of DCS as it is right now. I truly wish ED would put allot more work into the ground objects AI and damage model. Oh and for the OP, if a DCS level tank was released it would be a first day buy for me. Hell even if a FC3 level tank was released it would be a must have.

[Terrorvogel]

Of course yes...

[QuiGon]

I bought SB, I bought DCS, sure I would buy a SB-like DCS module!

[Pigmachine]

I have SB PRO PE (the last 3 or 4 versions) and love it, but don't play it so much, as I don't feel 'useful' enough to take part in multiplayer battles, and single player is not really that fun in the long run.

But I'd love to see it incorporated here too, even with the same pricing as the original (if we who already own it would get it a bit cheaper ;) )

 

But the SB have solid trees and no x-ray ai-tanks, max speed for when having your TOW raised etc etc. So maybe when edge comes out it would be different.

[BitMaster]

Has anybody of you thought about the Cluster you will need t run that ?

 

I mean, hey, this won't be peanuts but serious crunching in the TFLOPs arena.

 

You ask for many cores and even more memory and a very good connection

cause this scenario wont make sense with 14 players, even with 32 players you will feel isolated. 64-128 players is more what is needed to fill those maps with live. You need the BF/CoD kind of ground action and that needs at least 32 players to make fun, sometimes even 64 players are somewhat alone on large maps, at least in some BF mods. So say you have about 50 troops on the ground, infantry and armor and you need another 50 airborne on another server.... hmmm I suggest we move to ubuntu server cluster right away and forget about Windows cause nobody will be able to pay the server fees you will face renting that MS license Monster, Linux is way cheaper in renting per month and offers superb cluster support, which you will need, or you have a very good scaling SMP and 2 Xeons with a high Core count, PCIe based Flash Storage and enough RAM. It could work on 1 server, no doubt, just no server that only has 4 DIMM slots, 1U and a sits in a Datacentre for 99$/month. That wont do it.

If you go the multiple machine way you need a fast interconnect, 10Gbit which has a significant lower latency that 1Gbit....and 10x more ooomps.

With a DAS Flash drive in each you can copy GByte's in a fraction of a second.

 

I am afraid anything below this military grade Workstation/Server setup will not be stutter or lag free when you calculaze that many time stamp relevant data. You can see DCS like a video stream, anything that doesnt come in time is lost or worse, cause a delay or stutter.

 

Get that stream greased well !

 

That machine wont sell under 10k, rather 20-30k if you want it all in one :)

 

 

Who knows, in 5 years this power is available in any modern Desktop PC and by then ED coud be there too, so it is not that bad looking at all, if ya live long enough :)

 

 

Bit

Edited November 13, 2014 by BitMaster