When will there fourth generation il2?

[ITAF_Rani]

2 weeks !!?

[Irishratticus72]

Community : "When will we see a new gfx engine for Il2 GB"?

 

Janson: "Bear with me now....." 

[sevenless]

14 hours ago, FlyinCoffin said:

When will there fourth generation il2?

 

About 2030. Be sure...

[Rjel]

19 minutes ago, Irishratticus72 said:

Community : "When will we see a new gfx engine for Il2 GB"?

 

Janson: "Bear with me now....." 

Not sure who Janson is but if that’s what you see flying a GB game, you might want to crank up the graphics settings a notch or two. 

[Robli]

34 minutes ago, Rjel said:

Not sure who Janson is

 

William "Wes" Janson.

Edited November 18, 2021 by Robli

[Irishratticus72]

2 hours ago, Rjel said:

Not sure who Janson is but if that’s what you see flying a GB game, you might want to crank up the graphics settings a notch or two. 

That there is pure 4k!

[simfan2015]

3 hours ago, IckyATLAS said:

incredible 3D demos of an engine like Unity 2021.2

If you look at what a.o. DCS is trying to do with using the Vulkan API and reading how difficult that seems to be (rewrite graphics pipeline, shaders ...) then I would imagine using a completely DIFFERENT engine like Unity instead of those used by IL-2 GB, CloD, DCS ... is almost impossible.

What I do think I know about Unity is that it also as/had(?) quite some limitations (in the past) and that it would not be the best option for large worlds like those needed for ... flight sims. 

But indeed those Unity (2021.2) DEMOs do like real nice. 

Edited November 18, 2021 by simfan2015

[WheelwrightPL]

3 hours ago, IckyATLAS said:

Maybe it is more economically viable to do everything oneself as you can master and tweak everything as you please, but it is difficult to be expert on everything.

 

 

 

Exactly, the devs should concentrate on what they do best: creating believable WW2 aircraft models, and leave the world generation to specialized engines like Unreal 5 or Unity. If they choose to keep their geriatric engine on life support indefinitely, the have absolutely zero chance to catch up. And the gap will only get worse every year, until casual players who are interested in graphics abandon the sim, and only relatively few dieharts remain. This situation will make it very hard to sustain the business. On the other hand a quick adoption of 3rd part graphics engine will open the floodgates of WarThunder players who appreciate graphics but itch for something more realistic and engrossing.

[Rjel]

31 minutes ago, Irishratticus72 said:

That there is pure 4k!

Really? It looks pre 2K to me. 

[AEthelraedUnraed]

3 hours ago, Soilworker said:

One thing that bugs me about what is done on the current engine (however I don't know if it's a limitation or merely a design choice) is there is nothing metallic in the cockpits - by that I mean reflective (not including glass).

Everything's a tradeoff. They could make reflective surfaces, but they'd be much more expensive to the GPU.

 

3 minutes ago, WheelwrightPL said:

Exactly, the devs should concentrate on what they do best: creating believable WW2 aircraft models, and leave the world generation to specialized engines like Unreal 5 or Unity.

Unreal and Unity are both excellent general game engines for teams that don't have the resources to write their own engine. But being very generalised engines also means that they excel in preciously little. If you write your own engine, you can optimise it for your specific use case which always gives better results than what Unreal and Unity can do - provided you've got the necessary budget and knowledge.

[[F.Circus]FrangibleCover]

Between DVD, dynamic markings and new clouds there's a lot of investment in the core components of this version of Il-2 right now, so I don't think we're going to be getting a clean break new game any time soon. They wouldn't be doing all of this if they're just going to release Normandy and call it a day.

[ShamrockOneFive]

4 minutes ago, AEthelraedUnraed said:

Unreal and Unity are both excellent general game engines for teams that don't have the resources to write their own engine. But being very generalised engines also means that they excel in preciously little. If you write your own engine, you can optimise it for your specific use case which always gives better results than what Unreal and Unity can do - provided you've got the necessary budget and knowledge.

 

Exactly. My experience with Unreal engine being used in a sim is Train Sim World where the developers still had to write an entire sim engine that is underpinned by Unreal. It works but it does have some severe limitations that they have slowly chipped away at.

 

I think it's good that the folks at 1CGS built their own engine over a decade ago and have been able to keep it competitive all this time. They continue to punch well above their weight with it.

[Soilworker]

1 minute ago, AEthelraedUnraed said:

Everything's a tradeoff. They could make reflective surfaces, but they'd be much more expensive to the GPU.

 

 

I don't agree with the "much more" statement. Of course it will require some more graphics processing power, sure but not a lot.

 

We have reflective planes as I mentioned before such as the P-47 & P-51 to name just two and I didn't hear of anyone having to buy a new GPU to use them.

 

Same for the glass in the cockpits and those reflections are a lot more detailed and complex than what would be needed for metal surfaces (in my slightly educated opinion). 

 

Reflection maps have been around for a long time at this point and have run fine in many demanding games on much slower GPUs over the years. 

[WheelwrightPL]

12 minutes ago, AEthelraedUnraed said:

Unreal and Unity are both excellent general game engines for teams that don't have the resources to write their own engine. But being very generalised engines also means that they excel in preciously little. If you write your own engine, you can optimise it for your specific use case which always gives better results than what Unreal and Unity can do - provided you've got the necessary budget and knowledge.

 

Sorry, but I don't buy it, some of the demos for those engines show view distances far in excess of what IL2 engine is capable of producing. For example: look at the tree lines in IL2 engine and how they get cutoff, despite those trees not even being fully 3d rendered (that's why they twirl so unrealistically at certain angles). And then look at the vistas that those other engines can produce and not even break a sweat.

[BlitzPig_EL]

Games built on Unreal, etc... look great while you are trudging across their landscape on foot slaying dragons and taking arrows to the knee.  However, flying over them at 15,000 ft running 300+ mph, you will find them a lot smaller than you think, and their limitations will become quite glaring.

[AEthelraedUnraed]

27 minutes ago, Soilworker said:

I don't agree with the "much more" statement. Of course it will require some more graphics processing power, sure but not a lot.

 

We have reflective planes as I mentioned before such as the P-47 & P-51 to name just two and I didn't hear of anyone having to buy a new GPU to use them.

 

Same for the glass in the cockpits and those reflections are a lot more detailed and complex than what would be needed for metal surfaces (in my slightly educated opinion). 

 

Reflection maps have been around for a long time at this point and have run fine in many demanding games on much slower GPUs over the years. 

At the very least, it requires a couple of extra (non-linear=slow) function calls and an extra texture lookup. That's for the most basic not-very-realistic reflections. Double to quadruple that for better reflections with multiple blended probes.

 

31 minutes ago, WheelwrightPL said:

Sorry, but I don't buy it, some of the demos for those engines show view distances far in excess of what IL2 engine is capable of producing. For example: look at the tree lines in IL2 engine and how they get cutoff, despite those trees not even being fully 3d rendered (that's why they twirl so unrealistically at certain angles). And then look at the vistas that those other engines can produce and not even break a sweat.

Sorry, but this shows that you know little about how game engines work. Showing a couple trees in Unity from a large view distance is one thing, showing literally millions of trees in a 40km radius along with entire cities and atmospheric effects while running VR is something else entirely. Just the LOD lookups would eat your FPS. And that's just the graphics part. If you've ever worked with Unity's physics engine, you know how fussy it can be. Look at it the wrong way and your 10000kg He-111 is on its way to the moon within the blink of an eye. That's because it's made primarily to deal with relatively low forces, such as the ones you'd see in an FPS or RPG.

 

I'm not saying you couldn't make these things work in Unity or Unreal, eventually, but it'd require so much time and resources to implement it without screwing up other engine components that you don't have control over, that it frankly isn't worth the effort and you're much better off writing your own engine.

[PatrickAWlson]

Rebuilding from the ground up is not really how software works these days.  Games used to be tens or hundreds of thousands of lines of code.  You can rebuild that.  Now they are millions.  You don't walk away from that kind of investment.  

 

If your code is good then you don't have to rewrite.  You focus on upgrading one element at a time based on priorities.  If your code is not good (i.e. not neatly separated) then it is best to do that first without even attempting to change functionality.  This is referred to as addressing tech debt.  Then you change it from there .

 

The only way that you throw away a million plus line code base and start from scratch is if it is rotten and dated in every respect.   If you are in that state you are already in deep trouble.

[WheelwrightPL]

11 minutes ago, AEthelraedUnraed said:

Showing a couple trees in Unity from a large view distance is one thing, showing literally millions of trees in a 40km radius along with entire cities and atmospheric effects while running VR is something else entirely.

 

 

Did you look-up "nanites" in Unreal 5 ? They are designed exactly to address use cases such as: "million of trees". Here is the quote from https://docs.unrealengine.com/5.0/en-US/RenderingFeatures/Nanite/ : "Authoring content for Nanite is no different than traditional meshes except that Nanite can handle orders of magnitude more triangles and instances than is possible for traditionally rendered geometry".

[oc2209]

16 hours ago, BraveSirRobin said:

Sorry, but PTO needs carriers.  

 

Jason has clarified that the chief consideration in making 4 engine bombers is not technical limitation, but rather more trouble than it's worth from a development/economic standpoint.

 

Operating from that consideration, why would anyone think that developing several carriers, battleships, cruisers, and destroyers (to have an actual battle group and not just a carrier by itself), would be economically feasible or desirable for the current dev team? Above and beyond the inherent difficulties of making certain Japanese ships that likely wouldn't have adequate technical data to accurately recreate them in the first place.

 

That's asking for a huge time/effort/money investment from the developers, just so we can experience the thrills of carrier ops.

 

I'm saying it's not worth it. Not if the price is so high that the devs decide to never make any PTO modules at all, just because they're under the impression that no one wants the Pacific without carrier ops.

 

The solution is simple. When faced with serious impediments, unconventional thinking is required. Unconventional thinking in this case would be finding ways to make Pacific maps that can include diverse plane sets (all major US Navy planes, and the most common Japanese Navy and Army planes) without having ridiculously long flying distances between islands, and without involving carriers.

 

It's possible. As long as we don't rigidly adhere to prohibitive standards of what's absolutely necessary, it's all quite possible.

[AEthelraedUnraed]

5 hours ago, WheelwrightPL said:

Did you look-up "nanites" in Unreal 5 ? They are designed exactly to address use cases such as: "million of trees". Here is the quote from https://docs.unrealengine.com/5.0/en-US/RenderingFeatures/Nanite/ : "Authoring content for Nanite is no different than traditional meshes except that Nanite can handle orders of magnitude more triangles and instances than is possible for traditionally rendered geometry".

...yet when looking at them from a distance still a magnitude less efficient than the simple solution the Devs came up with for trees. The strength of nanites lies in high-poly objects. We don't need high-poly trees as we won't see them up close (although, I must admit, for Tank Crew some better foliage would be appreciated).

 

Again, I'm not saying that it isn't possible to make a good sim in Unity or Unreal. I'm saying that they are good frameworks for the average generic game. However, IL2 is not generic but *very* specialised (both large and very short viewing distances, millions of objects, complex weather, complex physics with high speeds, weights and power). I'm sure you can tweak Unity/Unreal and build workarounds to make it work acceptably well (which takes a *lot* of effort as well, of which what ShamrockOneFive posted about Train Sim World is an excellent example), but in the end you depend on software other people wrote and you have little control over. If Unity decides to launch a new version with a feature you need, you may well find that it breaks much of what you tweaked and you have to spend weeks or months redoing everything just to get this one new feature (this is not hypothetical, this actually happens. I'm sure it's not hard to find real-world examples if you google a bit. I know at least a couple). If you have control over all the code, you can always redo a bit of the base code to add in the new feature while having to redo as little as possible.

 

I'm also not saying that the Devs should do *everything* themselves in the unlikely case that they'd decide to start anew. There's a thousand good ready-to-use solutions on the web, and as a Dutch saying goes, "beter goed gejat dan slecht bedacht" (it's better to steal a good idea than come up with something bad). Every good programmer will agree. I just do not think either Unity or Unreal are one of those good solutions as they are simply too complete packages and it'd take too much effort to apply them to the IL2 use case.

Edited November 18, 2021 by AEthelraedUnraed

[BraveSirRobin]

1 hour ago, oc2209 said:

 

Jason has clarified that the chief consideration in making 4 engine bombers is not technical limitation, but rather more trouble than it's worth from a development/economic standpoint.

I’m sure that making B-17s isn’t an issue at all.  The problem is the resulting slideshow when you encounter even a small formation of bombers.  Just imagine what will happen when you encounter a carrier group when all the AAA opens fire.  The people developing this game aren’t morons.  They know what the issues are.

 

On the other side, the Pacific war without carriers has about as much commercial attraction as the Spanish civil war.  Which is to say, none. 
 

So they need to figure out the AI issues relative to big ships if they want to do PTO.

Edited November 18, 2021 by BraveSirRobin

[Gambit21]

1 hour ago, AEthelraedUnraed said:

...yet when looking at them from a distance still a magnitude less efficient than the simple solution the Devs came up with for trees. The strength of nanites lies in high-poly objects. We don't need high-poly trees as we won't see them up close (although, I must admit, for Tank Crew some better foliage would be appreciated).

 

Again, I'm not saying that it isn't possible to make a good sim in Unity or Unreal. I'm saying that they are good frameworks for the average generic game. However, IL2 is not generic but *very* specialised (both large and very short viewing distances, millions of objects, complex weather, complex physics with high speeds, weights and power). I'm sure you can tweak Unity/Unreal and build workarounds to make it work acceptably well (which takes a *lot* of effort as well, of which what ShamrockOneFive posted about Train Sim World is an excellent example), but in the end you depend on software other people wrote and you have little control over. If Unity decides to launch a new version with a feature you need, you may well find that it breaks much of what you tweaked and you have to spend weeks or months redoing everything just to get this one new feature (this is not hypothetical, this actually happens. I'm sure it's not hard to find real-world examples if you google a bit. I know at least a couple). If you have control over all the code, you can always redo a bit of the base code to add in the new feature while having to redo as little as possible.

 

I'm also not saying that the Devs should do *everything* themselves in the unlikely case that they'd decide to start anew. There's a thousand good ready-to-use solutions on the web, and as a Dutch saying goes, "beter goed gejat dan slecht bedacht" (it's better to steal a good idea than come up with something bad). Every good programmer will agree. I just do not think either Unity or Unreal are one of those good solutions as they are simply too complete packages and it'd take too much effort to apply them to the IL2 use case.

 

Unreal ended up being a stellar choice for Assetto Corsa Competizione - they physics is outstanding and it looks beautiful.

I wish I had time to drive it. I'm not certain to what extent the engine was modified.

[WheelwrightPL]

1 hour ago, AEthelraedUnraed said:

...yet when looking at them from a distance still a magnitude less efficient than the simple solution the Devs came up with for trees. The strength of nanites lies in high-poly objects. We don't need high-poly trees as we won't see them up close (although, I must admit, for Tank Crew some better foliage would be appreciated).

 

This guy tested nanites and achieved 600 billion triangles at playable framerates, that's more than sufficient to render real-looking trees at high fidelity:

 

 

[oc2209]

49 minutes ago, BraveSirRobin said:

On the other side, the Pacific war without carriers has about as much commercial attraction as the Spanish civil war.  Which is to say, none.

 

Right. Comparing the legendary Zero, Oscar, George, Wildcat, Hellcat, and Corsair, to low-performance pea-shooting planes (many of which are early models of planes we can already fly), in a prelude to WWII instead of WWII itself, is really an accurate comparison.

 

A great deal of combat in the Pacific occurred without carrier actions. Very few battles were strictly carrier versus carrier. The vast majority were carrier vs land-based units, land vs land, or land vs carrier.

 

Since kamikaze missions aren't going to be a career option, we can forget about land vs carrier.

 

Beyond that, think about this for a moment: how will the career structure work for brief battles like Midway or the Coral Sea? They can't. Are the devs going to make an entire Midway map just for multiplayer, with an incredibly short career mode therein? And further, are the devs going to build entire battle groups and carrier task forces (baby flat tops, British carriers with steel decks, etc, all necessary eventually once we go down the 'full naval experience' rabbit hole) without having a lengthy career mode in which to use them?

 

The brevity of carrier vs carrier engagements means that, to maintain the current standard of careers, we can only really make careers out of land vs land and carrier vs land operations.

 

Something like, oh, I don't know, Guadalcanal and Burma would totally satisfy these parameters.

 

Let's go on to the end game scenario, though. The situation in which we can fly the best Japan's got, versus the best America's got. This is a 1945 scenario. This would entail massive (impossible to implement in a game kind of massive, like D-Day armadas) American task forces launching aircraft at Japanese land targets. 

 

There can be no real Japanese attacks on the carriers, besides kamikaze. There is no legitimate carrier-carrier combat by this stage of the war. Therefore, we're going to need fully-rendered and operational carrier groups for only the American side, and only then, to serve as floating landing strips for naval planes to return to after attacking whatever Japanese land targets.

 

These carrier groups will serve no function other than to land on. I want to emphasize this point. Doesn't that seem like a vast waste of resources, to you?

 

So these are the options as I see them:

 

Make carrier battlegroups for Japan and the US, and have them duke it out for 3-5 days as per the usual length of carrier battles. With typically only 1 of those days comprising the majority of combat.

 

Or make US carrier battlegroups that don't actually engage in any battles, but more or less sit off the coast of mainland Japan or Okinawa/Iwo/whatever island, to pound Japanese land forces.

 

In both cases, there's a huge amount of development involved for very limited payoff. There's the excitement and novelty of landing on a moving carrier deck. That's it. Or there's the excitement of attacking carrier groups at sea for all of a day's length.

[BraveSirRobin]

4 minutes ago, oc2209 said:

 

Right. Comparing the legendary Zero, Oscar, George, Wildcat, Hellcat, and Corsair

Legendary for carrier ops.  That’s the point.  

[BlitzPig_EL]

New Guinea is the answer.  Mostly an Army show for both sides.  A multi year career possible for both sides, with many possibilities for online play.

 

Do it.

[AEthelraedUnraed]

4 minutes ago, WheelwrightPL said:

This guy tested nanites and achieved 600 billion triangles at playable framerates, that's more than sufficient to render real-looking trees at high fidelity:

Not sure what you want to show with this. Nowhere do I state that nanites isn't a rather exciting new technology. Neither do I state that nanites wouldn't be able to render a forest (in fact, I even state that nanites would be useful for Tank Crew). All I state is that what IL2 does (at least at large distances - which often is the case in flight sims) is optimised for the specific use case of IL2 and as such should be faster than nanites *for that specific use case*.

[oc2209]

1 minute ago, BraveSirRobin said:

Legendary for carrier ops.  That’s the point.  

 

No, it's not.

 

The Corsair was mainly land-based, at least early on. Zero didn't need and wasn't exclusive to carriers. Japanese Army planes didn't need carriers. The most famous/powerful late-war Japanese naval planes barely touched a carrier deck if ever.

[BraveSirRobin]

1 minute ago, oc2209 said:

 

No, it's not.

 

The Corsair was mainly land-based, at least early on. Zero didn't need and wasn't exclusive to carriers. Japanese Army planes didn't need carriers. The most famous/powerful late-war Japanese naval planes barely touched a carrier deck if ever.

The most famous Japanese aircraft barely touched a carrier deck?  
 

lol

 

That’s ridiculous nonsense.  The ONLY famous Japanese aircraft is the Zero, and it’s famous for carrier ops.  

[oc2209]

1 minute ago, BraveSirRobin said:

The most famous Japanese aircraft barely touched a carrier deck?  
 

lol

 

That’s ridiculous nonsense.  The ONLY famous Japanese aircraft is the Zero, and it’s famous for carrier ops.  

 

Read more carefully.

 

I said the most famous late-war Japanese aircraft.

 

And no, the Zero isn't the only famous one. Anybody who has even a passing interest in the Pacific knows about the George. The oft-overblown legends of it taking on dozens of Hellcats and surviving, etc.

[Gambit21]

2 minutes ago, BraveSirRobin said:

The most famous Japanese aircraft barely touched a carrier deck?  
 

lol

 

That’s ridiculous nonsense.  The ONLY famous Japanese aircraft is the Zero, and it’s famous for carrier ops.  

 

Have you read Guadalcanal by Richard B Frank?

I think you'll like it.  

 

Read that, and I think your opinion on the Zero and what makes it famous will at least adjust a bit.

No argument that carrier ops will have to come, but there's much more to the war than carriers. In fact for 18 months after Watchtower both sides had their carriers stashed out of harms way. Thus the fiercest, and longest-term, "force parity"  fighting was land based. Guadalcanal/Solomons, New Guinea.

 

That said, yes carriers will be important to a PTO product at some juncture. This can be immediately, or later,  but make no mistake there's room for 2 entire releases at least that don't even need carriers.

There's more than one way to skin the PTO cat. 

 

 

 

 

[BraveSirRobin]

1 minute ago, Gambit21 said:

 

Have you read Guadalcanal by Richard B Frank?

I think you'll like it.  

 

Read that, and I think your opinion on the Zero and what makes it famous will at least adjust a bit.

No argument that carrier ops will have to come, but there's much more to the war than carriers. In fact for 18 months after Watchtower both sides had their carriers stashed out of harms way. Thus the fiercest, and longest-term, "force parity"  fighting was land based. Guadalcanal/Solomons, New Guinea.

 

That said, yes carriers will be important to a PTO product at some juncture. This can be immediately, or later,  but make no mistake there's room for 2 entire releases at least that don't even need carriers.

There's more than one way to skin the PTO cat. 

 

 

 

 

I’m fully aware of the history of all these aircraft.  But I’m not the target audience.  Neither are you.  They’ve already got us.  For non WW2 dorks the PTO is carriers.

7 minutes ago, oc2209 said:

 

Read more carefully.

 

I said the most famous late-war Japanese aircraft.

There aren’t any.  The fact that you can’t grasp that is the problem.

[Gambit21]

2 minutes ago, BraveSirRobin said:

I’m fully aware of the history of all these aircraft.  But I’m not the target audience.  Neither are you.  They’ve already got us.  For non WW2 dorks the PTO is carriers.

 

I can't argue with that.

 

However on that note, I think carriers can be an good "carrot" to hold back, and bring in even more customers for a second PTO release.

Lots of options really - no shortage on that score with the Pacific.

 

[BraveSirRobin]

1 minute ago, Gambit21 said:

 

I can't argue with that.

 

However on that note, I think carriers can be an good "carrot" to hold back, and bring in even more customers for a second PTO release.

Lots of options really - no shortage on that score with the Pacific.

 

I’m not sure that this dev team can afford carrots.  They seem to be one failure away from screwed.  If we get a non carrier PTO module, it better be because they already know that they can do carriers,

[ACG_Bobo]

14 hours ago, Zooropa_Fly said:

 

Have you never been to Aberdeen ?

I had a buddy from Aberdeen, he had a had a buddy who had a girlfriend all his mates called 1860.  She looked like she was 18 from behind, but from the front, well you get the idea.
No offense meant to anyone, it's just funny.

[BraveSirRobin]

11 minutes ago, Gambit21 said:

 

I can't argue with that.

 

However on that note, I think carriers can be an good "carrot" to hold back, and bring in even more customers for a second PTO release.

Lots of options really - no shortage on that score with the Pacific.

 

By the way, they have realized the importance of iconic hardware up until now.  We got a flyable Ju52 for Stalingrad because duh.  We’re getting a flyable C-47 because duh.  
 

And if they can figure out a way to make carriers work… B-17…. 

[Gambit21]

4 minutes ago, BraveSirRobin said:

By the way, they have realized the importance of iconic hardware up until now.  We got a flyable Ju52 for Stalingrad because duh.  We’re getting a flyable C-47 because duh.  
 

And if they can figure out a way to make carriers work… B-17…. 

 

I think the transports were stellar decisions. 

I want a float Ju-52

[Avimimus]

1 minute ago, Gambit21 said:

I think the transports were stellar decisions. 

I want a float Ju-52

 

A float anything for that matter... hopefully the tech will be refreshed for it *if* we get FC.IV

 

I'd love an Arado Ar-196 artillery spotter (to go with an Fi-156 and Fw-189)... artillery spotting observers are a lot more 'action oriented' when you realise that an artillery battery has more firepower than any aircraft of the war - a little U-2VS with a radio has more firepower than anything else built!

[BraveSirRobin]

5 minutes ago, Gambit21 said:

 

I think the transports were stellar decisions. 

I want a float Ju-52

I spend the summer in Greenville Maine.  They’re currently restoring a float C47 to flyable condition so that it can be sold.  Should I let them know that you are interested?

[Gambit21]

11 minutes ago, Avimimus said:

 

A float anything for that matter...

 

Exactly