Engine upgrade and DX12 coming to ETS2 and ATS!

[flight50]

^Please use a translater.

[Berniyh]

Windows 7 and 8 are dead. There is just no denying that. If you're still on that, you need an exit strategy desperately.
Actually, you needed that years ago.

If you don't like Windows 10/11 (like me), then there are still Linux and OS X as alternatives.
If those are not acceptable to you either, then you're screwed, with or without ETS2/ATS.

[lelux]

The image on the last blog... except the truck... something new is arriving in the graphic compartment? What do you think?

[Some newbie driver]

One of the main pro DX12 arguments back then was HLSL since conversion tools back then were not mature, but that should've improved since then.

On a more broader topic, people need to understand that game developers create a lot of code tools of their own tailored for their own needs and partially linked the way their games are built (including the very basic fact of which engine is used). And those tools are NOT a joke, not at all; when properly done they save HUGE amounts of time to developers; so they mean HUGE savings in money but also less mistakes, more reliable schedules in roadmap milestones, less stress on the staff... Proper tools an documentation to face a migration from one API to another is the peak of the iceberg of the whole tool set. An important part indeed, but just a part of the whole.

That's one of the reasons every time somebody come here asking a "new engine for the game", theya re just showing how little they know about real developing. Just because X,Y or Z game they like are labeled as using "engine A, B or C"; that doesn't mean that the change will be any good for this game; starting by the fact that devs should throw to the bin a big part of all those tools they had been developing internally for decades.

And about what you mentioned, searching internet about what's better in that kind of API migrations; in my experience is the kind of searches that produce a huge pile of crappy results of random webpages full of "filling articles" too swallow at the best and really misleading at the worst. One need to have certain knowledge to sort the good from the bad on those searches and to know exactly the kind of words to use on the search engine to try to produce some useful results. And even so, too often I'm finding myself lurking beyond the 4th page of search engines (Google but others too) in order to find something useful for my needs. In IT topics internet is really full of absolute garbage.

Regards

PS: @lelux if you mean the fog, MAYBE, just MAYBE (with a ton of grains of salt) is a tease of a better fog implementation in the game. But it's more probable just an after-effect applied to the image to made it more teasing. It wouldn't had been the first time if it would be so. So far take only what's clearly teased, another new truck. Something that it's not b ad on itself; so better not be greedy wanting anymore of that announcement to avoid further disappointment.

[Tomsun]

@Some newbie driver The problem is not if it is difficult to make the new game engine (it is not just changing one switch and compile again), its the problem how long this topic is discussed (i remember writing first questions regarding this in my FX-8320 times)

In the past higher and higher clock speeds were curing this issue, but since the time of 4C/8T cpus as crown of genesis are HISTORY and Corecount/Threadcount is EVERYTHING even with lower clockspeeds today.

Since this change of paradigm there is only ONE way to go and that is towards multithreading OR accepting the limitations of it own Gamengine (that means in fact handcuffs for the mapdesigners and featureengineers).
The GE can't bare more and in fact some areas are for the current GE overengineered. Of course a multithreaded GE is not levelling equally, but perhaps it would be possible to extract some systemic tasks to unload the Gamenegine for the meantime until we get the big throw.

[Berniyh]

And about what you mentioned, searching internet about what's better in that kind of API migrations; in my experience is the kind of searches that produce a huge pile of crappy results of random webpages full of "filling articles" too swallow at the best and really misleading at the worst. One need to have certain knowledge to sort the good from the bad on those searches and to know exactly the kind of words to use on the search engine to try to produce some useful results. And even so, too often I'm finding myself lurking beyond the 4th page of search engines (Google but others too) in order to find something useful for my needs. In IT topics internet is really full of absolute garbage.

It's actually worse in this case. Because there are a zillion comparisons of DX11 vs DX12 vs Vulkan out there … from a user's point of view, i.e. benchmarks.
So a plain search won't really work, because it'll be clattered with all of those in addition to what you mentioned.

Luckily, I do code myself (but not in the graphics area), so I had an idea how and where to look and to some degree judge whether an answer is coming from somebody who really knows what he's talking about or not.
But even with that, there wasn't that much to be found and what you can find is mostly very old stemming from the time when both Vulkan and DX12 were still quite new.

[Someone Mysterious]

One thing that I’ve been wondering about, is how they plan on increasing the draw distance without breaking the entire map; because if the draw distance gets exponentially increased, won’t there be visible voids all over the map due to a lack of terrain in the previously unseeable parts of the map? That is of course if they decide to increase the draw distance. I hope I explained it well enough.

[Some newbie driver]

@Tomsun it take YEARS to make that kind of change, SEVERAL YEARS. It could take them less of course... at the cost to almost completely halt any other development. If we consider DX9 kick to be the first visible step (for us) to multi-core (they had to remove that old API and its limitations in order to start updating everything); I guessed back then 4 to 5 years more of wait if they had to keep releasing features in the meantime (and we had no little amount of code features since then).

That's how things are.

[flight50]

Draw distance talk seems to have piped down a ton since 2019 when Washington released featuring baked in models. SCS started baking in LOD models to avoid the popping in of objects. Its the popping that annoyed most people. With the past couple of ATS dlc's, I honestly don't see the need to increase the draw distance any longer. Butttt, this is just my opinion. I'd rather resources get applied elsewhere than render more draw distance.

DX12 Ultimate actually has a feature that I read up on recently in which resources are applied to close up things that are in focus and objects further away get less detail/resolution.

[Tomsun]

@Some newbie driver Yes i know it take years to do the changes. BUT Pavel is talking about this topic also for years and it would be interesting to know how far the development already took place. I only advocate for a kind of self-restraint in the mean time to keep it simable/playable.

There are a lot of ideas how to expand this Sim from the developers and from the users perspective (e.g. seasons, better ai, better ...) but the minimum requirement for this is the discussed GE and DX12 topic.

Yes i also advocat for a higher viewing distance to avoid the popping but currently i think this is not managable withe the current architecture.

If you compare e.g. Hungary , slavakia, UK to areas like austria rebuild its a difference like night and day, but you also recognize its the end of the possible spectrum currently.

The other question is if perhaps DX 19 is the actual standard what will be the reaction if then SCS launches DX12 (so what is the last timestanp DX12 is the big throw and not the introduction of an already oldstyle technology)
Another question is it possible to deload the Gamingthread of tasks in putting them in separate threads (e.g. autosave, navigation, ai) as a middlestep? Thats my amateur point of view, what would be logically. But to be honest i don't know if this approach would be possible. If yes the amount of ai for example would have no impact on the gameplay as long the ai thread is not overloaded. If just 15-20% of the load could be shifted to separate cores it would be a big achievement. My lack of understanding is is it just a manpower-problem (SCS is not a company like MS), or is it just a problem of technology what it makes it so hard or even impossible. And to be honest i can also live with an answer its not possible, we won't get it in the future, because it needs a rebuild from scratch (ETS3/ATS2). Or with an answer we planned for 2022 but due to unforeseen events its postponed to 2024 or even completely canceled.
From the side of SCS there is a big silence since the last statements from Pavel a long time ago.
The question how long will the build up of suspense last, and when will the community erode (remember trucksim is a niche product)
It could alsoo be a strategy to lift all parts of the map to a niveau of austria rebuild, as the last managable level of complexity. And just to do minor changes to the GE.
By the way is it possible to disable Autosave? Most of my stutters are related to that issue. If the framerates sometimes go down 10 FPS is not that big issue, as it usually gently wents up and down.