Possibility to use game engine and openbve

Hello

As we know openbve is left behind in terms of graphic with other trains games like trainz series or even like other opensource game like railsim.

I know there is benefit with making our own renderer like better integration with the code, less dependencies for library etc.


But better use the existing one isnt it ?

so therefore i suggest use game engine especially the renderer part, the game engine i see may compatible is godot.

Why? because godot already have pretty good GL3 and GL2 renderer and can even export to other platform like mobile even browser.

But yeah this easier said than done because there some stuff need to do like:

1. Upgrading openbve to use .net 6 because godot mono/c# version want net 6.

2. Cross languange code, because godot have their own language gdscript(python like) and openbve use mostly c# is may a bit tricky

3. Adapt openbve feature to godot


Also there some stuff i know that can make openbve significantly better when combine with godot:

1. Better graphic(not like big AAA game but better than current one)

2. Faster itteration when making UI stuff (i have a prototype from previous UI concept that i made but not done yet)

3. Keyframe based animation (i dont say current functional based bad but not intuitive, especially for people that not good at math and coding)

4. Possible an ingame route editor maybe ??

It's already been a talk somewhere, but the main thing is that just easier to create a new, fresh new game instead.

Doing almost all mentioned probably just saying that BVE not "BVE" anymore, making most older contents very hard in compatibility. At least that's what i thought.

IMO a game engine don't automatically make stuff look more beautiful.
It's true that BVE lacks support for features like light mapping, normal mapping, shadow, PBR etc., while a game engine with a more complete graphics pipeline can provide those features. However, it's still up to the content creators to set up the materials to utilize the functionalities, so the advantages won't apply to existing contents. It'll still look very plain if an engine such as Unity is used, but everything is all left with default settings.
And some features such as shadow are usually pre-baked, meaning that the map must be somehow compiled by the engine during map development for the effect to work, which won't apply to existing contents, plus that'll suggest a very different content creation process (Like becoming one must install Unity to make maps).
Postprocessing can be another factor in the look though, and there are some tools like ReShade that can apply some filters to the visuals of any game (including OpenBVE), to make some blooming effect etc.

There are definitely ways to make openBVE look almost realistic quality without relying on the graphics engine such as this example: https://www.youtube.com/watch?v=hIoil4eMMNE. However, it requires a lot more work on the developer end and I think that is where the problem is. It would definitely make the production of addons much easier and less time consuming if we did not have to put as much effort. 

In a way, I respect openBVE and its primitive graphics. It shows the varying levels of quality that are offered by different developers. But it can get tiring I think and as highlighted in the first post, there is a lot more that could be added to make things look even better.

I have heard some discussions about the possible implementation of Vulkan for openBVE maybe this is something to consider?

SP1900 wrote:However, it requires a lot more work on the developer end and I think that is where the problem is. It would definitely make the production of addons much easier and less time consuming if we did not have to put as much effort. 

That's the point I was making about pre-baked effects. It includes shadow, reflection, and ambient occlusion lighting etc. Many game engines have the function to automatically calculate lighting etc. and bake respective effects into the textures. The same effect can be theoretically achieved with preliminary means, just with more manual effort.
Though, it must be pre-calculated using such automated tool by developers (such calculation are too expensive to do in real-time). This would bring major change to the content making workflow of BVE, and I'd doubt if many creators are willing to take these extra setups, maybe they would just think it too much of a hassle. Plus, the function will require dedicated content to be made for it, and minor visual improvement can be made on existing contents with the approaches.

So I stated that using an engine won't magically make it look better. For example, Hmmsim Metro looks awesome, but that's more about Jeminie hiring professional modellers full-time to make all these detailed textures and lighting data, and less about it using Unreal.
Plus, support for light map and shadow map textures etc. can be added to OpenBVE's rendering system with not very much effort, but it's just about if content creators are willing to utilize such function (making these data for their model with some tool that requires learning)

Vulcan is a low-level topic, and it's more about optimization and less about visual improvement. If OpenBVE is to switch to Vulcan, it'll look exactly the same, just maybe with some more FPS.

Honestly, more FPS sounds perfect to me. Sure, the graphics engine is very limited and outdated by this point, but the capability to handle more complex models with high FPS is more than welcome.

Rakago wrote:Honestly, more FPS sounds perfect to me. Sure, the graphics engine is very limited and outdated by this point, but the capability to handle more complex models with high FPS is more than welcome.

Currently OpenBVE has an "old" GL2 renderer and a "new" GL3 renderer, IIRC there's an option to enable the new renderer. The new renderer should be pretty efficient already, albeit as limited as the old one.
The Vulcan Graphics API is often considered to be designed better (designed later so more experience to learn from) and hence better optimized, while the earlier developed OpenGL (3) has been and still is widely used, so they're both quite up to date by current standards.
If Vulcan support were to be added, the current OpenGL renderer would still be there as a fallback, since many older devices don't support Vulcan.

zbx1425 wrote:Plus, support for light map and shadow map textures etc. can be added to OpenBVE's rendering system with not very much effort, but it's just about if content creators are willing to utilize such function (making these data for their model with some tool that requires learning)

Personally, I would not be opposed to the creation of a new function or file format, anything to make the creation of high quality addons much easier but it is important to make sure that it does not completely render the previous work completely useless or that it ruins any additional functionality. 

If a new file format type is introduced to utilize this new function and graphical upgrades, we have to think about how easy it is to learn and use. And also, will it require developers to have to update all their addons to a new format? How easy will it be to convert addons created in the legacy format to be able to use the new functionality? Because the old style of addon creation has been used for years, many developers will not immediately upgrade to use new features unless the new features are relatively easy to use and provide some sort of improvement or shortcut to make addon creation easier.