Status of the game engine (Prism 3D)

[GreenMan36]

It's been 23 years since the first game on this engine, 12 years since this game launched on this engine.

Looking at the graphics of the game, its noticable the engine is getting old and the performance it's giving is sub-par for what we come to expect of modern games and their engines.

I am aware the engine has changed over the last 23/12 years, updated, upgraded, improved. I've read there are plans to rewrite/rework core parts of the engine. I wonder how that's going and what parts exactly. I don't think a few pipeline updates for newer rendering API's and graphical effects will be enough. Over the years the gaming industry has gone full speed ahead with amazing optimization techniques, allowing for very impressive games like MSFS.

I feel like more drastic changes might need to be made, I don't see ETS2 going for another 10 years like this.

[garyh2462]

Not everyone can afford a high end PC. With big graphic changes, come problems for people with low end computers. This is you're opinion and you are entitled to it. But for me, the graphics are fine. They can be enhanced with all sorts of mods. Progress takes time. Companies prioritise what they feel is important for continued customer satisfaction, I would say, they had kept the right balance. Write to SCS and see what they say. All the best. Gaztop

[Wolfgang]

1.50 will be the rework of the rendering engine, specifically the move to DX12. Multicore is also in the works.

Please be aware that SCS wants to also cater to the lower end spectrum of the gaming hardware and not exclusively to the high to highest end gaming hardware spectrum.
Please also keep in mind that SCS is very far away from being any AAA company like Microsoft (FS2020) with Billions of Dollars at their disposal yearly and with a ton of backend technology available.

And before anyone goes with "engine too old": That means that also Unreal Engine is too old as it is 26 years old by now and should not be used at all.

[Capt. Carrot]

The upgrade was originally intended/hoped to be released at the back end of last year but this has slipped as was alluded to in the christmas livestream.

[GreenMan36]

Thanks for the updates.

I am not exactly saying the engine needs high-end graphics; I wouldn't be able to run it, but for what it is offering today, modern engines offer the same for much better performance. I play a wide range of games from AAA to games that barely get 100 players.

As for the rework, I'm hyped for it, but skeptical of its improvements. Who knows? I feel like they're not transparent enough on the progress, so that's mainly why I'm asking and sharing concerns.

And yes, Unreal is old. But it's also completely different from what it used to be. The Unreal Engine from 1998 should not be used, just like Unreal (the game) is not getting any more updates. Gold Source evolved to various versions of the Source Engine and later into Source 2. Each iteration brought significant changes, additions, refactors, and complete rewrites. The same happened to the Unreal Engine. Also, Zouna, the engine for FS2020, Monopoly Plus, Ratatouille, etc., dates back to 2004 (actually 1990).

The differences are that these engines all had significant changes and updates quite consistently over time. Which ETS doesn't seem to have had. On top of that, they were designed and maintained for a very wide range of games, from platformers to simulators and shooters. So their architectures are battle-tested and robust.

Being a programmer, working on old codebases that haven't been changed much except for bloating, I am all too aware of how beneficial it sometimes is to not continue working on what you have but to remake parts or the whole from scratch. So yes, I am inclined to call the engine old and rigid. Not that it may not be salvaged, but that I believe it might be easier and provide a better outcome if majorly rebuilt, instead of reworked and patched up.

I'm mostly just concerned about transparency and that the engine will continue falling behind in performance and capability in favor of compatibility.

[Wolfgang]

What makes you think that Prism3D didn't have any significant changes? Is it just guessing or do you have any kind of source code to back it up?

Since Prism3D is not a commercially available engine SCS doesn't need to put a number behind it and be "Look everyone! More shiny!" because it won't do anything.

Also, don't get me wrong but I call BS on you being a programmer. We had to many "programmers" in here that talk so much garbage which can be proven wrong by a simple Google search.

@snowymoon (the person who made the TAA mod) describes it best:
https://forum.scssoft.com/viewtopic.php?p=1893863#p1893863

Originally posted by snowymoon:

I wrote this message because I see many people talking about "engine is old and there are limitations" etc, there isn't any limitation. This is not a physical engine. Yes, It becomes harder when game engine is complex but Its always possible to add any new features unless you need like DX12 or Vulkan, and Its always a lot easier to do it inside game engine, because if It was harder developers would do that outside of engine but they don't because thats not the case at all.

[mobsky]

I wish they would focus on the term "simulator" more than worrying about shiny fancy "graphics", there is nothing "simulator" about this. This is just a video game about driving a truck from a to b. There is no "simulation". Nothing will ever happen to your truck and so on.

I wish a big company buy's this or make an actual "simulator" instead of focusing on "dlc's" which they call a "country" that is made of some cities with 2 roads....

[GreenMan36]

What makes me think that? From the way the game, looks and runs to the layout of the files and how VR is implemented. Its like a codesmell, and it doesn't smell right. And as you said, they're not backed by MS with a ton of money to invest in technology.

And no it doesn't need a big number behind it but semantic versioning should for sure be used. And even just sharing that number every once in a while would do a lot for the transparency I was talking about.

As for me not being a programmer, you can check my GitHub under the same name ^-^
Most work I do is in private repo's or for companies but hopefully its enough for you to trust my word ;)

And for the "Its always possible to add any new features" part... yeah... sure...? But there's ways to do things "right" and a way to get things "done". Anyone can program, but few can make a good program.

I can't see the software architechture and I'm not coming in here like "Broo its been 10 years where is ETS3?", "This game engine is so bad, they should move to unreal." or any of that bs.

I'm coming in, explaining the engine is factually old, I heard there were some changes planned, sharing some hope and vision. Then asking what are the changes, how are they coming along and will that be enough for the next 10 years of ETS2?

The game is heavily focussed on assets and DLC, not much on simulation or technical aspects. The most impressive part of the engine is the input devices they support. But things like modding, simulations, graphics, performance and engine features are lacking.

I love picking up ETS from time to time to drive for hours and forget about my day, but

[Frate]

Obvisously there have been a lot of changes to the engine since they first used it 23 years ago.
Compare any SCS title from 20 years ago to the newest version of ETS2. You don't even have to go that far back, just compare ETS2 V 1.0 with 1.49, if you wouldn't know, you would think it's a completly different game.

Instead of renaming the engine for every new version, Unreal 2,3, etc, they kept the same name, but that doesn't mean there hasn't been any significant changes.

ETS2 used to be 32bit dx9, now it's 64bit dx11 and soon it will be dx12. If you are really a programmer, you should now that porting a game engine to a new API, isn't as easy as just changing the number.

[Wolfgang]

Originally posted by GreenMan36:

What makes me think that? From the way the game, looks and runs to the layout of the files and how VR is implemented. Its like a codesmell, and it doesn't smell right. And as you said, they're not backed by MS with a ton of money to invest in technology.

VR is only being worked on by one person in their free time. VR is not officially supported as per Steam Store page. So taking it as being in any way representative is like saying that the steriotypes of a country are representive of the people living there. It isn't.

Originally posted by GreenMan36:

And no it doesn't need a big number behind it but semantic versioning should for sure be used. And even just sharing that number every once in a while would do a lot for the transparency I was talking about.

For what? No one needs that. There is not going to be any company wanting to licence it so ti doesn't make any sense.

Originally posted by GreenMan36:

As for me not being a programmer, you can check my GitHub under the same name ^-^
Most work I do is in private repo's or for companies but hopefully its enough for you to trust my word ;)

Do not care.

Originally posted by GreenMan36:

And for the "Its always possible to add any new features" part... yeah... sure...? But there's ways to do things "right" and a way to get things "done". Anyone can program, but few can make a good program.

You clearly haven't seen the work snowymoon did. What they did is not only a first but also insanely impressive.

Originally posted by GreenMan36:

I can't see the software architechture and I'm not coming in here like "Broo its been 10 years where is ETS3?", "This game engine is so bad, they should move to unreal." or any of that bs.

I'm coming in, explaining the engine is factually old, I heard there were some changes planned, sharing some hope and vision. Then asking what are the changes, how are they coming along and will that be enough for the next 10 years of ETS2?

To quote a dev:

Originally posted by Max (SCS dev):

Nobody can tell you anything about it. Those who know aren't allowed to talk and those who talk know nothing.

Originally posted by GreenMan36:

IThe game is heavily focussed on assets and DLC, not much on simulation or technical aspects. The most impressive part of the engine is the input devices they support. But things like modding, simulations, graphics, performance and engine features are lacking.

I love picking up ETS from time to time to drive for hours and forget about my day, but

Modding is lacking? With all due respect, but are you high? Just look in the workshop and make a basic Google search of "ETS2 mods" and tell me it is lacking. There is basically nothing you can NOT mod. Saying that it is lacking is akin to saying that a billionair is poor.
And you said yourself:

Originally posted by GreenMan36:

I am not exactly saying the engine needs high-end graphics; I wouldn't be able to run it

That tells me that your setup wouldn't be able to run better stuff anyway.
The vanilla game performance is great so if you have issues there it is either your hardware and/or you using mods.

However, if you expect next-gen graphics then this won't happen as SCS is also catering to those with lower end gaming hardware (as I said above already and which is shown in the minimum specs needed for the game). Wanting to have photorealism leaves those people out.
Not to mention that great graphics are not making a good game.

[GreenMan36]

Originally posted by Frate:

Obvisously there have been a lot of changes to the engine since they first used it 23 years ago.
Compare any SCS title from 20 years ago to the newest version of ETS2. You don't even have to go that far back, just compare ETS2 V 1.0 with 1.49, if you wouldn't know, you would think it's a completly different game.

Instead of renaming the engine for every new version, Unreal 2,3, etc, they kept the same name, but that doesn't mean there hasn't been any significant changes.

ETS2 used to be 32bit dx9, now it's 64bit dx11 and soon it will be dx12. If you are really a programmer, you should now that porting a game engine to a new API, isn't as easy as just changing the number.

1. You're missing the point.
2. Sure buddy, I'm not a programmer, gl with that.
3. 32x/64x is a frkn compiler switch and not even relevant to the discussion.
4. Porting an engine to a different API isn't easy, I never said it was just changing a number, but its quite doable if programmed for it from the get-go. Also upgrading to a newer version of DX is easy compared to alternatives. Vulkan is a pain.
5. I haven't said there are no significant changes.
6. I specifically mentioned 12 years for a reason.

[GreenMan36]

Originally posted by Wolfgang:

Some big ass quote because Steam, god I hate this

Alright, didn't know only one person worked on VR but that anologue misses a bit hard lmao.
Semantic versioning is an industry staple, you'll always need some way of versioning any type of software you have and semantic versioning makes it incredibly clear when something's happened that affects more than just what changed. Not having semantic versioning is a reason to not work at a company. Even now, when we are talking to each other, I cannot compare the current engine to anything before as its tied to the game and not the technology.
And no I did not see the work that snowymoon did, I'm sure its impressive, the whole game is.
Having things internal is exactly why I'm concerned, asking for any info and asking for transparancy.
Yes modding is lacking, all I am seeing is some asset mods and a few tweaks. I haven't seen anyone make significant changes to the game other than different models, textures, places and roads.
As for my setup, its on my profile. I can run VR and AAA (2022) games just fine, the performance of the ETS engine (Prism3D) is just lacking.|

I didn't come for an argument, I came for clarification and sharing concerns. You seem to have shared what you can, so I welcome someone else to chime in with some new info that hasn't been mentioned, if any.

[Un Pollo]

Originally posted by Wolfgang:

1.50 will be the rework of the rendering engine, specifically the move to DX12. Multicore is also in the works.

Please be aware that SCS wants to also cater to the lower end spectrum of the gaming hardware and not exclusively to the high to highest end gaming hardware spectrum.
Please also keep in mind that SCS is very far away from being any AAA company like Microsoft (FS2020) with Billions of Dollars at their disposal yearly and with a ton of backend technology available.

And before anyone goes with "engine too old": That means that also Unreal Engine is too old as it is 26 years old by now and should not be used at all.

Man, i'm agree with you but...pls, don't be silly.
Unreal engine 5 it's a complete new engine (same as the u3 and u4 and so on). Every new version of the engine was a complete new engine...so UE5 it's not a 26 years old engine...maybe you are a little bit confused.
The fact that it calls unreal engine doesn't mean that is always the same engine of the 1998.
You cannot compare the two things.
However, their engine IS TOO OLD, and that's a fact.
BUT...SCS, at least, it's finally implementing a new version of the core engine, with DX12 and Vulkan support...so, even if they would make the game more modern with new effects and lighting and texture, etc. i don't think that it would be an "high end" restricted game...they are not stupid, and they know that the majority of the players of this game have a low end pc, i think...so, yes, people shouldn't worry about it.

[GreenMan36]

Originally posted by Un Pollo:

Man, i'm agree with you but...pls, don't be silly.
Unreal engine 5 it's a complete new engine (same as the u3 and u4 and so on). Every new version of the engine was a complete new engine...so UE5 it's not a 26 years old engine...maybe you are a little bit confused.

No they do build on top of each version, they call it a new version once they drop a bunch of new features on top of the old one.
They make a major release, work on it for years and then work on the next major release in tandum.
Unreal engine 5 was in development allongside unreal engine 4 for 2 years before being announced. And UE5 was built on top of UE4. Its the same engine with major overhauls for each major release.

[Wolfgang]

Originally posted by Un Pollo:

Originally posted by Wolfgang:

1.50 will be the rework of the rendering engine, specifically the move to DX12. Multicore is also in the works.

Please be aware that SCS wants to also cater to the lower end spectrum of the gaming hardware and not exclusively to the high to highest end gaming hardware spectrum.
Please also keep in mind that SCS is very far away from being any AAA company like Microsoft (FS2020) with Billions of Dollars at their disposal yearly and with a ton of backend technology available.

And before anyone goes with "engine too old": That means that also Unreal Engine is too old as it is 26 years old by now and should not be used at all.

Man, i'm agree with you but...pls, don't be silly.
Unreal engine 5 it's a complete new engine (same as the u3 and u4 and so on). Every new version of the engine was a complete new engine...so UE5 it's not a 26 years old engine...maybe you are a little bit confused.
The fact that it calls unreal engine doesn't mean that is always the same engine of the 1998.
You cannot compare the two things.
However, their engine IS TOO OLD, and that's a fact.
BUT...SCS, at least, it's finally implementing a new version of the core engine, with DX12 and Vulkan support...so, even if they would make the game more modern with new effects and lighting and texture, etc. i don't think that it would be an "high end" restricted game...they are not stupid, and they know that the majority of the players of this game have a low end pc, i think...so, yes, people shouldn't worry about it.

Seems like you only partially understood the post. Saying that Prism3D is too old is like saying Unreal Engine is too old. In both cases you assume that no work has been done on either engine. Which is not the case and ignoring all the stuff that has been done in all these years.

[Wolfgang]

Originally posted by GreenMan36:

Originally posted by Wolfgang:

Some big ass quote because Steam, god I hate this

Alright, didn't know only one person worked on VR but that anologue misses a bit hard lmao.
Semantic versioning is an industry staple, you'll always need some way of versioning any type of software you have and semantic versioning makes it incredibly clear when something's happened that affects more than just what changed. Not having semantic versioning is a reason to not work at a company. Even now, when we are talking to each other, I cannot compare the current engine to anything before as its tied to the game and not the technology.
And no I did not see the work that snowymoon did, I'm sure its impressive, the whole game is.
Having things internal is exactly why I'm concerned, asking for any info and asking for transparancy.
Yes modding is lacking, all I am seeing is some asset mods and a few tweaks. I haven't seen anyone make significant changes to the game other than different models, textures, places and roads.
As for my setup, its on my profile. I can run VR and AAA (2022) games just fine, the performance of the ETS engine (Prism3D) is just lacking.|

I didn't come for an argument, I came for clarification and sharing concerns. You seem to have shared what you can, so I welcome someone else to chime in with some new info that hasn't been mentioned, if any.

How does semantic versioning apply to software that is NOT commercially available? It is a thing for commercial products and Prism3D is NOT one of these. You can not buy it as it is the in-house engine of SCS and not from some third party company. They are also not selling it to anyone. Stating any version number to the public is useless because not needed.
You should also understand that SCS is not really keen on talking about stuff that isn't finished yet because broken promises (even if not promised in the first place) are a thing and upset a lot of people.
You also really need to see the modding that has been done, there are several economy mods, cargo mods, map mods, truck and trailer mods and whatnot. If you find anything lacking then it is either because it is too complicated to do or too niche for anyone to have done it. And since you are a alleged programmer it should be easy for you to do it. If someone can make a TAA mod in 2 months then modding something yourself should be super easy.

You also might want to read my first post again, specifically the part on multicore being in the works.
I heavily implies that the game is currently very CPU bound (in fact pretty much pegging one core at 100% all the time) thus limiting its current performance. However with a proper CPU that can be reduced.