Unreal Engine 5

[LucShep]

 

On 12/20/2021 at 11:47 PM, Eugel said:

A few years back, many people in the sim racing community were sceptical when "Assetto Corsa Competizione" was announced to use UE4, like "what, a serious simulation with the Unreal engine ??"
And it had/has? terrible performance in VR compared to other sims and at least in the beginning had a lot of other VR related problems as well (haven´t played it in a while, so can´t say how the current status is)

Just like DCS, ACC is another game built with absurd levels of detail (ahead of its time even) built by a very small team that, funny enough, also end-up listing very, very optimistic hardware requirements for what it really needs.
The performance hit you see in ACC (with UE4) versus the original AC (with Kunos' in-house graphics engine), or even its other rivals, has to do with developer choices, not really to the graphics-engine itself.

On 12/19/2021 at 12:30 AM, Taz1004 said:

UE5 cannot be used for DCS because the ever changing nature of DCS.  Unreal is for games that release and be done with.  Even updating UE4 project from older to newer version of UE4 was pain.  Updating a project from UE4 to UE5 is almost like starting over.  And major version update for Unreal happens every 8 years or so.

Business model of DCS is both its strength and weakness.  Continuing to support old modules means you can't start fresh like Unreal Engine can.  You still see remnants of ancient DCS code in the game today.  And it gets harder and harder to update the engine as the game gets bigger with more modules.  And it has to keep getting bigger to make profit and keep it going.

And then there's performance side with UE5.  All those beautiful graphics comes at a cost.  You most certainly need RTX series for Lumen and Nanite.

 

FC3 used FC2 models/textures initially, in the early days of DCS World. They were updated later, weren't they?
The KA-50 BackShark, A-10C, P-51 Mustang and other modules have been updated. Assets will continue to be updated and created, always, regardless of game engine.

I don't buy the "Continuing to support old modules means you can't start fresh like Unreal Engine can", and I'm not sure where are you basing at for that assumption.

Developers of DLC content for FSX make content for XP11, P3D and now even MSFS. 
They're all highly complex hardcore flight-sims with different graphics, sounds, and physics engines, yet the very same content is seen accross the different platforms from the 3rd party teams, just like ones we have for so many DCS modules.

How many years have been between different graphics engines (DCS1.2, DCS1.5, and DCS2.5, cue next one) and the eternal ups and downs with performance hit, even steeper hardware requirements, endless bug-fixes, and whatnot?

I agree that UE5 would be a very long, difficult, huge task and challenge for ED, but the continuous development business model of DCS will also require a total revamp of the graphics engine at some point. It's inevitable.

For the fellas dismissing the UE5 idea, think about this...
How many years have we been waiting just for Vulkan implementation on DCS now? ....five or six years, I believe?
What if, in the end, the gain results by Vulkan in DCS end up being too small? 
What is the next trick in the sleeve? And then how long for that next one?

There's another point by using a 3rd party engine like UE5, which is the tecnical support provided if required.
 
And no, you don't have to have an RTX series for Lumen and Nanite.
Their point is to optimize work flow and performance, for techniques that, as of today, are known to be rather taxing and complex to produce.

With Nanite you can increase geometry complexity (higher triangle and objects counts) as frame budgets are no longer constrained by polycounts, draw calls, and mesh memory usage (something that DCS can really benefit from). 
You can use high-poly detailing rather than baking detail into normal map textures (another thing that DCS can really benefit from - the current 2K, 4K, 8K and 32-bit normals are utterly hideous and excessive, bogging performance and devouring VRAM).
The Level of Detail (LOD) is automatically handled and no longer requires manual setup for individual mesh's LODs (just imagine the time savings with development alone!), with the big bonus of loss of quality being rare or non-existent, especially with LOD transitions, because of it.

With Lumen it's different, it's a "Global Illumination" feature (like Ray-Tracing) and a separate option that, IIRC, is used in different levels, depending on pre-defined (non custom) quality levels set in the game's graphics options.
For example, by default, with "Epic" level it's set for a 30 fps budget (8ms Global Illumination and Reflections), while under the "High" level it targets 60 fps. Lumen is disabled under "Low" and "Medium" levels. AFAIK, it can be made as scalable option, that can also be switched ON and OFF in options.
Nvidia GTX1070 / AMD Vega56 or better GPU for performance is prefered for Lumen (also the minimum requirement for DCS as of today), which is pretty darn fantastic, amazing really, considering that most high detail games of today using Ray-Tracing work only with Nvidia, and usually  require an RTX2080 or better for 60+fps@1080P with RT enabled.

Edited December 22, 2021 by LucShep

[Taz1004]

59 minutes ago, LucShep said:

I don't buy the "Continuing to support old modules means you can't start fresh like Unreal Engine can", and I'm not sure where are you basing at for that assumption.

Just personal experience of working with Unreal Engine since UDK.  What you're not realizing is that if you use Unreal, you're relying on Epic.  If they make a change in their engine version, you either have to stop updating or do a major overhaul every 4 weeks or so.  For example, I was working on a project with UE4 and they changed the sequencer.  Updating my project means having to redo every cinematics.  Sometimes they just get rid of blueprint node or codes I was using altogether.  I once had to redo entire menu UI because of it.  Every version update of UE4 had some sort of adjustment that needs to be done in order for my project to work whether it be graphics, animations, or even base code.  And most of the updated features had NOTHING to do with my project.  Primary customer of Unreal Engine is consoles and that's what most of their features are targeted at.  Which is also why Epic don't care about backward compatibility.  So if ED were to use UE5 and keep the engine up to date, they will be spending time on things that has nothing to do with flight sim.

So this leads to not updating the engine version at all.  And when that gets outdated, starting over.  Like most game development such as IL2.  This is completely different than ED implementing Vulkan themselves.  In that case, ED is in control.

 

59 minutes ago, LucShep said:

The KA-50 BackShark, A-10C, P-51 Mustang and other modules have been updated. Assets will continue to be updated and created, always, regardless of game engine.

It's not about assets or textures or content.  It's about backward compatibility.  Why UE5 isn't compatible with UE4 or UE3?  DCS engine is updated incrementally to ensure backward compatibility.  If they rely on Epic, that's not gonna happen because as I said above, when Epic updates their engine, they will break things.

Again, Unreal is great for games that you release and done.

 

59 minutes ago, LucShep said:

And no, you don't have to have an RTX series for Lumen and Nanite.

I have GTX 1080.  Nanite and Lumen are unplayable.  You can find examples of UE5 on GTX 1080 and 1650's.  Yes, Nanite is future.  Not present.  Not to mention they're not compatible with VR.

Edited December 22, 2021 by Taz1004

[LucShep]

1 hour ago, Taz1004 said:

Just personal experience of working with Unreal Engine since UDK.  What you're not realizing is that if you use Unreal, you're relying on Epic.  If they make a change in their engine version, you either have to stop updating or do a major overhaul every 4 weeks or so.  For example, I was working on a project with UE4 and they changed the sequencer.  Updating my project means having to redo every cinematics.  Sometimes they just get rid of blueprint node or codes I was using altogether.  Every version update of UE4 had some sort of adjustment that needs to be done in order for my project to work whether it be graphics, animations, or even base code.  And most of the updated features had nothing to do with my project.  Primary customer of Unreal Engine is consoles and that's what most of their features are targeted at.  And if ED were to keep the engine up to date, they will be spending time on things that has nothing to do with flight sim.

So this leads to not updating the engine version at all.  And when that gets outdated, starting over.  Like most game development such as IL2.  This is completely different than ED implementing Vulkan themselves.  In that case, ED is in control.

 

I understand what you mean, but we're talking UE5, not UE4. 
My area is sound-design but I keep in touch with collegues in the graphics area. Comments I see from them is that EPIC has noted the critics of developers with UE4 projects (not as smooth as before with UE3), similar to ones that you've pointed. Speculative only but, things are rather optimistical for UE5 in that regard.

1 hour ago, Taz1004 said:

And if ED were to keep the engine up to date, they will be spending time on things that has nothing to do with flight sim.

So this leads to not updating the engine version at all.  And when that gets outdated, starting over.  Like most game development such as IL2.  This is completely different than ED implementing Vulkan themselves.  In that case, ED is in control.

That's no different than what happens today with DCS. With the added drawbacks of huge performance penalty and even steeper hardware requirements with each graphical update (1.5 versus 2.5, versus 2.7, for example). We're now entering in the era of "unicorn RTX3080ti for DCS or get busted".

And while it's true that ED are in total control then, if you're using a problematic graphics-engine (and a rather unoptimized one at that) with roots dating back 20 years ago, then you're always stuck with endless fixing and revamping, regardless.
As I see it, it's a catch-22.

It's also rather different to what we see with the latest IL2 GB (BoX), for which a brand new (and very good) game engine was built from scratch (and how I wish ED had done back in the day for DCS World), after the lesson learned from the disaster with IL2 CloD of Maddox.
And on this last one, and like it (finally, with Blitz and Desert Wings), I hope we don't have to wait for a decade to see DCS optimized. 

1 hour ago, Taz1004 said:

I have GTX 1080.  Nanite and Lumen are unplayable.  You can find examples of UE5 on GTX 1080 and 1650's.  Yes, Nanite is future.  Not present.  Not to mention they're not compatible with VR.

Unplayable? Now that's odd. 
I was on discord with some friends (like you, graphics artists) and one was streaming to us while playing around with Nanite, evaluating its performance.
When he ran it in 3rd person template, and play tested it without Nanite it ran at about 45-60 fps (depending on how many onscreen), with Nanite it wouldn't drop below 120 fps.
Thats mind blowing performance. Undoubtfully, like the best thing since sliced bread from what we've seen that day, like, many degrees above traditional static (and dynamic) meshes geometry system.

I haven't seen real examples of Lumen other than the few home-made examples on the youtube, but those also seem really promising IMO.

As for VR, there's a lot of unanswered questions surrounding UE5 but VR support is definitely one of the main goals, so it's really just a matter of time.

Edited December 22, 2021 by LucShep

[Taz1004]

30 minutes ago, LucShep said:

As I see it, it's a catch-22.

I agree it is a problem which is why I said DCS model is both strength and weakness on previous post.  And I don't believe for a second Epic will improve backward compatibility in UE5 but even if that is true, it will still happen eventually when UE6 comes out.  ED will have to start over if they are to implement latest features of UE6.  Which by that time, I expect DCS to have a lot more modules.

It's like improving your home versus rebuilding your home.  Both have pros and cons but the problem with Unreal is the cycle.  You would be rebuilding every 8 years... based on previous release cycle.

Nanite basically trades textures with polygons.  So if it was scene with pure polygons, then Nanite would give you improvement.  But comparison should be made between low poly/high texture versus high poly/low texture.  Not just same high poly scene with Nanite on/off.  So newer RTX cards with faster core are much better at handling polygons.  Where older gen still relies on textures.

Edited December 22, 2021 by Taz1004

[LucShep]

6 hours ago, Taz1004 said:

I agree it is a problem which is why I said DCS model is both strength and weakness on previous post.  And I don't believe for a second Epic will improve backward compatibility in UE5 but even if that is true, it will still happen eventually when UE6 comes out.  ED will have to start over if they are to implement latest features of UE6.  Which by that time, I expect DCS to have a lot more modules.

It's like improving your home versus rebuilding your home.  Both have pros and cons but the problem with Unreal is the cycle.  You would be rebuilding every 8 years... based on previous release cycle.

Different ways of looking at things then. I read that and feel it like the analogy of "I'm not buying Hearblur's F-14 Tomcat because Razbam's F-15 StrikeEagle is coming sometime in next couple of years". Rather use something right now that is worth my time for what I want.

Sure, either way, there is no perfect solution. But I see far more long term benefits with a brand new graphics-engine, rather than staying with this one.
And, IMO, UE5 seems the perfect candidate. If UE6 comes out in eight years, that's a whole different matter.

We've seen nearly a decade of development of DCS World to understand that optimization is not really something to hope for (performance issues during its whole exhistence), so I'm not really trusting the current (old) engine for big benefits, by staying with it.

By the way things are going with DCS, numbers of users reverting to older/outdated versions of DCS will increase more and more (I'm one of them) just to keep using the content they invested in, no longer being able to run latest versions with decent performance. Which means, not having access to newest modules and updates for the ones that they paid for.
With the hunger for prettier graphics and effects, and how the hardware (GPU) market is today, ED is forgetting this aspect. It will hurt them in the end.
 

6 hours ago, Taz1004 said:

Nanite basically trades textures with polygons.  So if it was scene with pure polygons, then Nanite would give you improvement.  But comparison should be made between low poly/high texture versus high poly/low texture.  Not just same scene with Nanite on/off.  So newer graphic cards with faster GPU are much better at handling polygons.  Where older gen still relies on textures.

I'm more interested in the application of it, beyond the precious performance benefits,  also for the benefits of development (better, faster). 

AFAIK, UE5 Nanite tech claims to eliminate polygon budgets and enable developers to directly use their high-quality source models without having to bake a low-poly version and manually create LODs.
I think you can understand that this is a very big step forward, also for speeding-up work flow for new content by simplifying things (more modules, etc).

This means the certain types of geometry could potentially be entirely represented using only geometry and no textures that approximate surface detail. A good application of this could be high-quality 3D scanning, probably even more usefull if laser-scanning a vehicle, or aircraft, or terrain (etc), which is becoming more and more common.

Not having to downscale models, create LODs, make normal maps, bake lighting (which can literally take hours if the scene is large enough), or optimize literally anything aside from one's code would be amazing. It would save so much valuable time. That along with the fact that it looks fantastic, and runs great.

That is, if it lives up to its promises which, after what I've seen, it seems to.

Edited December 22, 2021 by LucShep

[Taz1004]

6 minutes ago, LucShep said:

AFAIK, UE5 Nanite tech claims to eliminate polygon budgets and enable developers to directly use their high-quality source models without having to bake a low-poly version and manually create LODs.

I'm aware of the benefit if Nanite which is why I said it's the future.  But not until RTX becomes the minimum requirement.

And we're not even looking at the cost.  5% royalty will apply to 3rd party DCS developers and even campaign creators.  And it's 5% gross revenue, not profit.

[LucShep]

5 minutes ago, Taz1004 said:

I'm aware of the benefit if Nanite which is why I said it's the future.  But not until RTX becomes the minimum requirement.

And we're not even looking at the cost.  5% royalty will apply to 3rd party DCS developers and even campaign creators.  And it's 5% gross revenue, not profit.

Honst question (because I'm not understanding what you mean) - why is the RTX the minimum requirement for Nanite?

And what is the minimum GPU requirement for DCS right now for latest update? (please don't tell me you're totally happy with your GTX1080 performance?!)

[Taz1004]

5 minutes ago, LucShep said:

Honst question (because I'm not understanding what you mean) - why is the RTX the minimum requirement for Nanite?

And what is the minimum GPU requirement for DCS right now for latest update? (please don't tell me you're totally happy with your GTX1080 performance?!)

Because as I said above, texture still is more efficient than polygons for GTX cards.

And no, I'm not perfectly happy with GTX1080 but with software improvements such as FSR/NIS, I'm actually quite content currently.  My DCS experience has improved 200% since a year ago.  Enough that I'm willing to wait longer for GPU price to drop.  I honestly don't understand why some say they can't handle the performance.  Yes there are some bugs like the cluster bomb frame loss bug but that's more of bug than engine issue which I'm sure will be fixed soon.

[LucShep]

12 minutes ago, Taz1004 said:

And no, I'm not perfectly happy with GTX1080 but with software improvements such as FSR/NIS, I'm actually quite content currently.  My DCS experience has improved 200% since a year ago.  Enough that I'm willing to wait longer for GPU price to drop.  I honestly don't understand why some say they can't handle the performance.  Yes there are some bugs like the cluster bomb frame loss bug but that's more of bug than engine issue which I'm sure will be fixed soon.

If you can share some advice/settings I'd really appreciate it!
I tried a program on Steam called Lossless Scaling ($3 only) that allows any GPU to use FSR, NIS, and integer scaling.
The problem for me is that neither NIS nor FSR produced decent visual quality, had to run the game in windowed mode and that means triple buffered V-Sync, which introduces a lot of input lag, and wasted CPU resources as Windows is running in the background, and above all, there's no VRR support (which I really want to keep using).
 

Edited December 22, 2021 by LucShep

[Taz1004]

25 minutes ago, LucShep said:

If you can share some advice/settings I'd really appreciate it!
I tried a program on Steam called Lossless Scaling ($3 only) that allows any GPU to use FSR, NIS, and integer scaling.
The problem for me is that neither NIS nor FSR produced decent visual quality, had to run the game in windowed mode and that means triple buffered V-Sync, which introduces a lot of input lag, and wasted CPU resources as Windows is running in the background, and above all, there's no VRR support (which I really want to keep using).

FSR/NIS probably wont help you much at your setting.  It has benefit in VR because you can set any resolution you want in VR.  And you can compensate the image loss with slightly higher resolution.  But in 2D, you're fixed with specific resolution so all you'll be doing is trade performance for quality.  And FSR is performance improvement at some quality loss.  It's not quality improvement.

But 4K gaming is still considered high end.  It was one of boasted feature of RTX 3080.

Edited December 22, 2021 by Taz1004

[Mr_sukebe]

I'm with Taz on the pros and cons on this, and have seen it with business software/services.

Build it yourself and you get what you want.

Buy it in has HUGE benefits of being able to spread the cost of development across all of the purchasers.  However, then you have the issue of getting it to do what you want, which means either "adapting" business processes to follow the software design, or hardcoded customisation to the application.

As a couple of examples, consider:

 - Star Citizen, which is built on the CryEngine, and as I understand it has already been through one "transition" to a newer version of the engine and has huge levels of customisation, which will never easily port across to a newer version

 - Elite Dangerous, which is built by themselves and over time is simply ever improving.  When they started using it for 1st person, it was apparently a pile of poo.  But I tried it recently and it seemed OK to me, so clearly it's been given some love.

All have pros and cons.

The reality is that the group most likely to understand the options/risks/costs/timescales are ED.  For all we know, maybe they will buy in an external "engine", which "might" be Unreal or even the FS2020 underpinnings.

Edited December 22, 2021 by Mr_sukebe

[BiBa]

9 hours ago, LucShep said:

 

Just like DCS, ACC is another game built with absurd levels of detail (ahead of its time even) built by a very small team that, funny enough, also end-up listing very, very optimistic hardware requirements for what it really needs.
The performance hit you see in ACC (with UE4) versus the original AC (with Kunos' in-house graphics engine), or even its other rivals, has to do with developer choices, not really to the graphics-engine itself.

FC3 used FC2 models/textures initially, in the early days of DCS World. They were updated later, weren't they?
The KA-50 BackShark, A-10C, P-51 Mustang and other modules have been updated. Assets will continue to be updated and created, always, regardless of game engine.

I don't buy the "Continuing to support old modules means you can't start fresh like Unreal Engine can", and I'm not sure where are you basing at for that assumption.

Developers of DLC content for FSX make content for XP11, P3D and now even MSFS. 
They're all highly complex hardcore flight-sims with different graphics, sounds, and physics engines, yet the very same content is seen accross the different platforms from the 3rd party teams, just like ones we have for so many DCS modules.

How many years have been between different graphics engines (DCS1.2, DCS1.5, and DCS2.5, cue next one) and the eternal ups and downs with performance hit, even steeper hardware requirements, endless bug-fixes, and whatnot?

I agree that UE5 would be a very long, difficult, huge task and challenge for ED, but the continuous development business model of DCS will also require a total revamp of the graphics engine at some point. It's inevitable.

For the fellas dismissing the UE5 idea, think about this...
How many years have we been waiting just for Vulkan implementation on DCS now? ....five or six years, I believe?
What if, in the end, the gain results by Vulkan in DCS end up being too small? 
What is the next trick in the sleeve? And then how long for that next one?

There's another point by using a 3rd party engine like UE5, which is the tecnical support provided if required.
 
And no, you don't have to have an RTX series for Lumen and Nanite.
Their point is to optimize work flow and performance, for techniques that, as of today, are known to be rather taxing and complex to produce.

With Nanite you can increase geometry complexity (higher triangle and objects counts) as frame budgets are no longer constrained by polycounts, draw calls, and mesh memory usage (something that DCS can really benefit from). 
You can use high-poly detailing rather than baking detail into normal map textures (another thing that DCS can really benefit from - the current 2K, 4K, 8K and 32-bit normals are utterly hideous and excessive, bogging performance and devouring VRAM).
The Level of Detail (LOD) is automatically handled and no longer requires manual setup for individual mesh's LODs (just imagine the time savings with development alone!), with the big bonus of loss of quality being rare or non-existent, especially with LOD transitions, because of it.

With Lumen it's different, it's a "Global Illumination" feature (like Ray-Tracing) and a separate option that, IIRC, is used in different levels, depending on pre-defined (non custom) quality levels set in the game's graphics options.
For example, by default, with "Epic" level it's set for a 30 fps budget (8ms Global Illumination and Reflections), while under the "High" level it targets 60 fps. Lumen is disabled under "Low" and "Medium" levels. AFAIK, it can be made as scalable option, that can also be switched ON and OFF in options.
Nvidia GTX1070 / AMD Vega56 or better GPU for performance is prefered for Lumen (also the minimum requirement for DCS as of today), which is pretty darn fantastic, amazing really, considering that most high detail games of today using Ray-Tracing work only with Nvidia, and usually  require an RTX2080 or better for 60+fps@1080P with RT enabled.

 

A very elaborated feedback!

Thanks...

 

On the other hand, talking about the CPU 1 core Engine DCS is using: Isn't it about time DCS gets its CPU an extension! 

With all the modules, assets, mods, etc, it's a miracle and a great achievement how with one core DCS manages to get the job done. I fear soon the DCS One-Piston-Engine will soon crash against the digital sound barrier and won't keep pace with the complexities of the future sixth-generation digital fighters.

[SkateZilla]

the Revamp is already underway with the move to Vulkan, and pretty much in the final stretch if you go by the last news article mentions.

Assets across multiple platforms isn't anywhere near easily done, it's not copy/paste nor is it export using exporter A then export using exporter B.
Where as w/ DCS 1.2->2.8 it's just a matter of updating plugin, updating textures etc, re-exporting, The container format hasnt changed, it just gets updated.

Going from XP to DCS to MSFS, every single one of those models need to have animations redone as well as collision models etc, and every sim uses it's own proprietary container format.

Unreal Engine 4 was released in 2014, and started development in 2003, that's 11 Years, as it was a entire re-code from scratch.
Unreal Engine 5 retained the UE4 Core, 2-3 years, while UE5 Engine is released, it's not feature complete, they are adding things in phases.
As I said in another thread, no studio will write an entirely new graphics/physics/complete engine from scratch with the level of features that make it a AAA Engine in a matter of months, or even the short side of quarter of a decade.

For as good as they make Nanite out to be, please direct your attention to the limitations, especially in the animation and morphing sections.
Same with Lumen,

I suspect a lot of people were wow'd by the Matrix Tech demo, but didn't actually read through the entire tech documents for developers.

Also with UE, 5% of Sales go back to Epic, further removing income from ED and partners, partners who have no say in the game/graphics engine used, so why switch the UE5 and Force Partners/3rd Parties to sell on a platform where 5% is given to Epic.

Users and fans that don't develop see the neat tech demos and press releases, but don't understand the financials and limitations involved, as they are not talked about in those aforementioned demos and press releases.

Vulkan, will remove the DX11 API Draw call overhead, and offer ED a Modern API with the ability to drop in a endless number of plugins easily through the Kronos developer community.

Something DirectX 11 doesnt have.

Edited December 24, 2021 by SkateZilla

[wilbur81]

I'm SO pleased/excited about the current state AND future of DCS! Any other fancy trailers (the most recent one looking like DCS with MSAA cranked up :-)) for the 'other' newest thing honestly don't impress me all that much. DCS isn't just pretty...she's got a LOT of depth and is getting deeper still. 

[3WA]

On 12/15/2021 at 3:23 PM, Mr_sukebe said:

Probably worth looking at the Unreal site, that states that U5 was designed for map sizes upto 21kmsq.

Actually, you can now enable Large Worlds, and it will allow maps to 88 million km square by enabling 64 bit double floating point coordinates.

Flight Sim ON!  Lol, maybe.  Have yet to see real proof of how much power, etc. it would take.

[3WA]

On 12/22/2021 at 5:03 AM, Mr_sukebe said:

 

Build it yourself and you get what you want.

 

The problem with really old engines like DCS is, are they capable of simulating flight sim / vehicles / fps over whole worlds at modern graphics, or have they reached their limits?

I think at that point, you should start considering buying in to a modern engine, and how much of the old material you would be able to port over, and how long it would take.

I've downloaded UE5 and it really is a step up from the old UE4, which I really wasn't that impressed with.

EDIT: EH!  Double post.  Sorry.

Edited July 27, 2022 by 3WA

[Mr_sukebe]

3WA>

Had forgotten about this thread.

The comment that I'd read about map size was written on the Unreal site.  However, I couldn't help but notice that NOR looked like it was going to use Unreal 5, and quite obviously from their brief clips, there appeared to happily exceed the suggested boundaries.  That gives me a good deal of hope for the future regarding options on game engines that could be used.