Future plans for rF2?

[stonec]

That's nothing new in software development. From operating systems to applications, every big software project has to deal with its legacy and it's not that hard with some planning. There's two ways to deal with it: 1. throw everything into the garbage can and start again; 2. introduce new features, deprecate old ones to give time for content to catch up and later remove them. I don't know which one S397 will be going to take but it will greatly impact their future.

And nobody does your first option in practice. A game physics engine is probably C++ code that has for example basic physics equations formulated. Why would those need a rewrite? Rewriting something like rF2 from scratch would probably be 5-10 years of work for a medium-size team.

The most rewrite I can see would be rF3 moving to an industry standard game engine like Unreal Engine. By using standard engines you eventually end up saving work hours, but the initial effort required is of course quite big. Even then there would be no reason to move away from pMotor (rF2 physics engine).

 

[Bernat]

And nobody does your first option in practice. A game physics engine is probably C++ code that has for example basic physics equations formulated. Why would those need a rewrite? Rewriting something like rF2 from scratch would probably be 5-10 years of work for a medium-size team.

The most rewrite I can see would be rF3 moving to an industry standard game engine like Unreal Engine. By using standard engines you eventually end up saving work hours, but the initial effort required is of course quite big. Even then there would be no reason to move away from pMotor (rF2 physics engine).

I mean redoing the content since you're supposedly going to change the engine so much that old content can't be made compatible in any way. This implies big changes to the engine too but I was only referring to remaking the content..

Because, and this takes us to another big issue, if you don't make a substantial update to the engine, users won't see the point on migrating to the new title. Loosing all content for only a few new features, better graphics, and a lot of new bugs?

If you're going to put out a new title, it has to offer something new and worthy of the loss of content. And we will see this with AC2 probably. ACC offered something different to AC, but AC2 should rather be a very big upgrade or offer something really new from what AC offered, or their users will see it as a fail, not an upgrade but a downgrade because of the loss of content.

A new title isn't any kind of magic that fixes everything, it might only change some problems for others which could be worse. But we can dream it's magic.

Unreal Engine isn't a good engine for racing titles. Choosing the incorrect tools for the work can be an endless waste of time.

 

[davehenrie]

Unreal Engine isn't a good engine for racing titles. Choosing the incorrect tools for the work can be an endless waste of time.

Apparently it is not a good engine for Moon Landings as well. The visualization program that Intuitive Machines used to guide it's lander to touchdown was run through UE5. And it came down too fast, broke a landing leg, and fell over. (Swamp Castle anyone?)

 

[hitm4k3r]

At this point I don't see it as a viable option to do anything else than what they are doing allready. They are betting on their best horse using their inhouse technology and people seem to appreciate it. Developing a new engine from scratch costs money and what would be the benefit of switching to UE or Unity for gfx? I have barely seen anyone complaining about gfx, quite contrary. There are quite alot of people who think that LMU is the best looking racing sim currently. I also think some people underestimate the amount of new features that are under the hood when they request features to be ported back to rF2. Just the amount of detail that went into the whole hybrid system for all the different cars, track specific settings, settings specific to LMDh compared to the LMH cars, how the dash elemets are linked to it and how - for me the most impressive thing - how the sound is linked to the physics is something that I don't see them doing for rF2. When the battery is full for the 499P you won't hear the energy recovery system recharging under braking and the rears will be locking up due to worse brake performance. And it's not just a simple on/off sound change but very dynamic. It all fits together in a very visceral way and I think that's what keeps people playing LMU. None of the other studios has gone that far and I just don't see S397 going back to implement features as they have done in LMU, so going all in with full-nerd-mode. Maybe they will use the technology developed for LMU very far down the line as a base for rF3. But that's very very far away.

 

[stonec]

I mean redoing the content since you're supposedly going to change the engine so much that old content can't be made compatible in any way. This implies big changes to the engine too but I was only referring to remaking the content..

Because, and this takes us to another big issue, if you don't make a substantial update to the engine, users won't see the point on migrating to the new title. Loosing all content for only a few new features, better graphics, and a lot of new bugs?

If you're going to put out a new title, it has to offer something new and worthy of the loss of content. And we will see this with AC2 probably. ACC offered something different to AC, but AC2 should rather be a very big upgrade or offer something really new from what AC offered, or their users will see it as a fail, not an upgrade but a downgrade because of the loss of content.

A new title isn't any kind of magic that fixes everything, it might only change some problems for others which could be worse. But we can dream it's magic.

Unreal Engine isn't a good engine for racing titles. Choosing the incorrect tools for the work can be an endless waste of time.

Yep. We will see. The problem is already evident though. AC was released 10 years ago and the fact that there is no successor yet points to the problem. rF2 developers did the same as AC did with ACC to create a stand-alone limited series title. For both sims their predecessor's success is the main hindrance in creating the next major version of their sim.

Ultimately game studios are forced to do something though if they want to keep the company running and create recurring revenue. Maybe AC sales still serve as recurring revenue due to the player count, but rF2 I would think will not forever into the future. It seems that Codemasters/EA and iRacing got it right after all. Either release a new sim every year or have people paying for a monthly subscription. Unless you do this you will have all your long-term customers using your product for free and not seeing any reason to move on from it either.

 

[hitm4k3r]

Yep. We will see. The problem is already evident though. AC was released 10 years ago and the fact that there is no successor yet points to the problem. rF2 developers did the same as AC did with ACC to create a stand-alone limited series title. For both sims their predecessor's success is the main hindrance in creating the next major version of their sim.

Ultimately game studios are forced to do something though if they want to keep the company running and create recurring revenue. Maybe AC sales still serve as recurring revenue due to the player count, but rF2 I would think will not forever into the future. It seems that Codemasters/EA and iRacing got it right after all. Either release a new sim every year or have people paying for a monthly subscription. Unless you do this you will have all your long-term customers using your product for free and not seeing any reason to move on from it either.

Good points. An important difference that I see with how LMU does compared to rF2 and how ACC did compared to AC is that LMU does alot better allready than rF2 and ACC (in the beginning) aswell. ACC had something like 300 player peak at the beginning. You just couldn't do any meaningfull stuff with it, it ran like a$$ and there was very limited content and features. I was very sceptical before release if people would keep playing LMU, but it just doesn't seem to ware off, or not very quickly. I kind of predicted it as the content is very variable and I suspect once BOP get's implemented properly, once the game get's more stable and once cars show their different strength and weaknesses, people have quite alot of stuff to do testing all the different combos. There are a few hundred hours of enjoyment with that small amount of content allready. I mean, people are coping with all those bugs and glitches for very obvious reasons. It's simply a great and fleshed out racing expeirence and will only get better at this point.

 

[Lazza]

At this point I don't see it as a viable option to do anything else than what they are doing allready. They are betting on their best horse using their inhouse technology and people seem to appreciate it. Developing a new engine from scratch costs money and what would be the benefit of switching to UE or Unity for gfx? I have barely seen anyone complaining about gfx, quite contrary. There are quite alot of people who think that LMU is the best looking racing sim currently. I also think some people underestimate the amount of new features that are under the hood when they request features to be ported back to rF2. Just the amount of detail that went into the whole hybrid system for all the different cars, track specific settings, settings specific to LMDh compared to the LMH cars, how the dash elemets are linked to it and how - for me the most impressive thing - how the sound is linked to the physics is something that I don't see them doing for rF2. When the battery is full for the 499P you won't hear the energy recovery system recharging under braking and the rears will be locking up due to worse brake performance. And it's not just a simple on/off sound change but very dynamic. It all fits together in a very visceral way and I think that's what keeps people playing LMU. None of the other studios has gone that far and I just don't see S397 going back to implement features as they have done in LMU, so going all in with full-nerd-mode. Maybe they will use the technology developed for LMU very far down the line as a base for rF3. But that's very very far away.

The sound engine in rF2 is already amazingly capable. Really the only thing stopping what you describe from LMU is the hybrid physics (and appropriate events being made available for the sound engine - I think there remain some gaps for the physics already in rF2). Even if there are core sound engine enhancements in LMU I would say the rF2 one is at least 80% matching it. A huge step up from the list of available sounds and basic parameters we had since rF1.

The biggest hurdle I see for rF2's future is the sheer scope of the project - many of the core issues are still in LMU, and assuming they get fixed over time it will be in the context of a continually changing codebase - changing away from the rF2 one. Replicating those fixes here could be like starting over, and DLC probably won't cover the costs. All this assumes LMU gets to a point where devs aren't needed full time there.

As much as we've been used to having rFactor as a jack of all trades, we may eventually have to settle for games featuring specific series instead. rF2 will become less relevant over time as its features fade in comparison; but, I don't see that happening in a major way for years. Certainly not tomorrow.

 

[TonyM]

How would you rate LMU in comparison to rF2 for:
- the sounds
- the graphics
- the AI

 

[hitm4k3r]

How would you rate LMU in comparison to rF2 for:
- the sounds
- the graphics
- the AI

Much better in those areas obviously. On the other hand you have to deal with a few Early Access hickups and the shortage of content. But the content that is there is ouf outstanding quality accross the board. Similar level like the 992 in terms of modeling with a few nice features added on top of it. I really dig the windscreen getting dirty. Tracks are all on a similar level like Sebring. Only Mount Fuji looks a bit lowres at this point.

 

[RaceNut]

Much better in those areas obviously. On the other hand you have to deal with a few Early Access hickups and the shortage of content. But the content that is there is ouf outstanding quality accross the board. Similar level like the 992 in terms of modeling with a few nice features added on top of it. I really dig the windscreen getting dirty. Tracks are all on a similar level like Sebring. Only Mount Fuji looks a bit lowres at this point.

I really don't mind the lack of content in this case. The experience is good enough to warrant spending time to get reacquainted with tracks I've not driven in quite a few years. So, I'm taking my time and one by one, learning / relearning the finer points of each track and car combination. By the time I have these down, there should be some new content (fingers crossed).

 

[UGM 133A]

I hope LMU's success allows S397 to then make other singularly focused games. If that's what's needed to get the best out of the engine and licensing for a particular vehicle type, then I don't see it being much different to the end user than paying for new DLC in rF2, you could always jump between the sims and still have effectively the same physics engine with car specific features that makes the most out of the engine.

This might not be a popular opinion, but I think if it's done well for each game, it would change people's minds.

 

[pilAUTO]

I hope LMU's success allows S397 to then make other singularly focused games. If that's what's needed to get the best out of the engine and licensing for a particular vehicle type, then I don't see it being much different to the end user than paying for new DLC in rF2, you could always jump between the sims and still have effectively the same physics engine with car specific features that makes the most out of the engine.

This might not be a popular opinion, but I think if it's done well for each game, it would change people's minds.

I understand the principle, and it's not stupid.

But this does not correspond to everyone's use, for example mine which I will describe below so that you understand :

Apart from use on one or two other circuits such as for example the Targa Florio, I only uses Nürburgring mainly in its combined version.

I mainly use two lists of opponents, which I took the time to do well so that there are mainly 4 different car speeds, around 40-50 different cars or more than rfactor 2 choose when loading the game, 89 ai that is to say the maximum authorized by the track on rFactor 2, obviously the complete day-night cycle with a random Time start, a semi-scripted weather and the random one with in particular cloud management so that it changes a maximum over time, and a qualification mode only (no race) in which very quickly the 89 ai are distributed perfectly over the 26 km due to perfectly adjusted AI parameters, a large number of AI, and above all 4 categories cars in terms of speed, for example cars like Clio etc, btcc etc, gt3 etc, hypercar etc.

I did not give the details but all the parameters of the game which I know perfectly well since I have been using it for 10 years now are set so that I can have the most beautiful immersion linked or nürburgring, and have the most fun.

It's a very specific use of the rfactor 2, but at least it's really personal and personalized and I'm really having the experience I dream of having in real life in terms of driving.

For this, I obviously need a nürburgring with these layouts, obviously need a magnificent and realistic day-night cycle, a changing and realistic weather, above all obviously a breathtaking driving realism this which is the case here, but the number of AIs as well as especially their very wide variety is very important.

Furthermore, I am not at all interested in simulations which offer a choice not of a small number of cars but a choice of restricted driving variety like it IS in LMU/ACC etc.

(I want to drive historic, all periods, any type of transmission, et cetera,..................).

I find it much too reductive in term of variety and driving variety interest (cars especially but tracks too) for LMU but of course it's personal, I'm not saying that it can't suit many many people.
It doesn't suit me.

And the rare times when I'm tired of the Nürburgring, I go in particular but not only to the Targa Florio, a magnificent circuit in a magnificent atmosphere with a interesting height difference, 72 km with around 10,000 turns and even more, with any type of car or almost it's just exciting.

So as you can see, LMU or a very restricted rfactor 2 derived simulation of this type has very little or even no interest for some people like me.

But again, it's not a matter of opinion, it's just a matter of user and usage.

It doesn't suit me and it doesn't meet my expectations.

Hence my desire to improve rfactor 2, or in the worst case even if it is not what I prefer, an rfactor 3 taking up the majority of its philosophy but obviously improving it significantly on a number of aspects.

 

[Marcel Offermans]

Apart from use on one or two other circuits such as for example the Targa Florio, I only uses Nürburgring mainly in its combined version.

Sounds like you would be interested in a Nürburgring simulation, and there might actually be a market for that, since this is such a popular track that hosts so many different types of races. One track, many cars. I think I heard Gamermuscle propose such an idea in one of his streams recently.

 

[pilAUTO]

Sounds like you would be interested in a Nürburgring simulation, and there might actually be a market for that, since this is such a popular track that hosts so many different types of races. One track, many cars. I think I heard Gamermuscle propose such an idea in one of his streams recently.

it is true that if motorsport game and s397 make a simulation based on rfactor 2 as is the case for LMU but especially the universe of a nürburgring as a whole, if it is done well, it could be incredibly good at least from the point of view of people like me who are fans of it.

The question is :

Simulation being already a niche hobby, truly ultra-realistic Rfactor 2-style simulation being even more niche, could it work on a commercial level even if it is true that the Nürburgring is one of the most popular circuits in the world if not perhaps even the most in reality I don't know?

In any case, I buy directly without asking myself any questions and I see afterwards.

However, for it to work for people like me, it would have to be ultra realistic and as it stands, that means that it necessarily uses the rFactor 2 physics engine.

(or something other than the physics engine of rfactor 2 only if in the next two or three years a studio manages to do as well as possible in this area, which still seems quite unlikely to me).

 

[Love Guitars n Cars]

On iRacing Nürburgring has it's own series which is successful so yes it's own title with many different cars would definitely sell.

 

[FERNANDO CABRERA]

Ok here are my takes on what I have seen:

1. LMU implemented the new interface in Java and when setting up an event it sends an http call lo a listening service in localhost with a JSON object with the session setup; you can inject an object to it with your own parameters. This will possibly be closed in the future.
2. The mod manager from rfactor2 runs… but lmu is validating the track hashes to wat seems like a hardcore base; if a mod is installed the game engine will not load the new content even if you try to spoof it or infect it to the game execution
3. Of course, all mas files in lmu are encrypted… and you can not back port the content to rfactor2; this is basically the same protection been use for the paid dlc.
4. There are references in the log files for a livery creator, but could not find how to connect to it.
5. Everyone knows that plugins from Rfactor 2 runs, even vr runs… so the underlying code base is largely share between the two products; but as some one pointed out, could be a fork because the Ai in lmu seems more evolve in some regards to the one in rf2, and also the wheels uis are miles better from what we had in rf2
6. Hybrid system has been build from ground to work how it do in wec and it will continue to evolve in lmu, not seen that been back ported any time soon to rf2.

Refactor 2 is my go to simulator, and would love to see thrive, would love lo have all official content from lmu and NASCAR game also in Rfactor2 as paid dlc... Would love to see more people in the daily races…
One can only hope…

 

[MikeV710]

Apparently it is not a good engine for Moon Landings as well. The visualization program that Intuitive Machines used to guide it's lander to touchdown was run through UE5. And it came down too fast, broke a landing leg, and fell over. (Swamp Castle anyone?)

Can you please give me a link to you´re source?

Thanks

PS.: I also think UE Is not a good engine for simulation.

 

[davehenrie]

Can you please give me a link to you´re source?

Thanks

PS.: I also think UE Is not a good engine for simulation.

I was just channel surfing about 2hours prior to the moon landing and they had the virtual display of the various guidance & radio directions, to fill time the announcer gave a brief rundown of the system and commented about it running on UE5. That is the best I can give you as a source.

 

[MikeV710]

I was just channel surfing about 2hours prior to the moon landing and they had the virtual display of the various guidance & radio directions, to fill time the announcer gave a brief rundown of the system and commented about it running on UE5. That is the best I can give you as a source.

Thanks, and no worries I appreciate you´re reply.

Cheers

 

[UGM 133A]

Thanks, and no worries I appreciate you´re reply.

Cheers

Like with any taxpayer funded program in the US, they typically try to maximize their bang for their buck by using COTS (commercial off the shelf) hardware/software, so using UE5 for simulations makes sense compared to making their own software for it. I worked on the hardware side of the Artemis program for some of the guidance stuff when I was still at Honeywell and I remember similar cost saving efforts.