Bring back (+ or - ) part of LMU improvements to rF2 (ONLY most profitable for users + MSG)

[Lgel]

In my view the main blocking problem to the evolution of RF2 has been the huge existing base (content, tracks, cars, official, and mods).

An evolution to LMU standards of RF2, would require modification of all existing base that you want to run with the improved RF2, which is a daunting task (may be only recent DLCs), or branching to different forks of code depending on old or new content, with a heavy impact on performance and a maintenance headache.

An other option would be to allow modding in LMU, leaving the weight of the adaptation to modders, and widening the appeal of LMU.

Cheers.

 

[Miguel Cardoso]

I understand the idea of creating a new game, but as a player, I don't quite agree with it. They should have done that within rFactor 2, as that would have improved rFactor 2 itself. Let's be clear, are we going to buy a game with 8 tracks and a series of cars? We'll easily get bored driving the same things over and over.

I understand MSG's position, but if it were iRacing that got the ACO license, they would hardly make a game just for Le Mans.

As for rFactor 2 itself, it's a shame there haven't been any updates in over a year. I think the developers should continue developing the game and not be so concerned about mods or tracks that will become obsolete. It should be up to the modders to update their content to comply with the new standards.

It's incredible that we're still running GT3 championships with cars that no longer race, like the Porsche 991 GT3R.

 

[Lazza]

I understand the idea of creating a new game, but as a player, I don't quite agree with it. They should have done that within rFactor 2, as that would have improved rFactor 2 itself. Let's be clear, are we going to buy a game with 8 tracks and a series of cars? We'll easily get bored driving the same things over and over.

Would and did. Many successful games have limited content in comparison to games like rF2, doesn't stop people buying them.

I understand MSG's position, but if it were iRacing that got the ACO license, they would hardly make a game just for Le Mans.

ACO might have gone with MSG to have a standalone game, rather than being some content in an existing one.


As for current cars - that's the point of buying a game based on a series. You don't buy F122 and then complain it doesn't have 2024 cars in it.

People will say they'd happily support the studio by buying cars and tracks, but apparently not when they're in a separate title.

 

[AMillward]

Would and did. Many successful games have limited content in comparison to games like rF2, doesn't stop people buying them

ACC shipped with just GT3 and whatever tracks were used that season it came out. People jumped on it in an instant.

but now they don’t need to beg for the Nordschleife anymore it’s gone a bit quiet XD

 

[pilAUTO]

So thank you all for giving your point of view.

I just read the fifteen messages that arrived in the last three days, there is just one thing I wanted to answer.

I have already given my opinion on the subject but I just wanted to repeat what I said without being really precise or being ultra-argued :

I have tested LMU all the same and it is rather good.

As I have already said, what really interests me is the realism of piloting in addition of course to content that interests me otherwise it is not worth it.

So that is the first thing I checked in a way, the realism of piloting.

It is quite at the level of rFactor 2.
If it really is better, it is really extremely little and marginal, which in itself would already be very good of course.

But to say that LMU is really a notch above in terms of physics, I absolutely disagree.

It does indeed have very specific features, particularly related to hypercars, but in terms of overall driving, frankly it is extremely close to rfactor 2, if there are improvements, they are very small and linked to the specificities of the cars and setups. Very small does not mean that it is useless, of course, since it is very, very far from the most important area, even a very small improvement is already fantastic.

Except that I think that many people are far too influenced by what the developers say, these are great new tires blah blah blah, I do not believe for a single second, for example, that the tire model has really progressed significantly.
Just see how the initial tire model of rFactor 2 has progressed extremely slowly over the years, I honestly do not believe it, and I think there is a placebo effect of sorts.

Then, very important point, I have been using rFactor 2 for 11 years and I am really an advanced user. The base of the 2 games being totally identical with of course a code that has drifted for LMU, in reality the vast majority of the settings if not almost all are exactly the same.

It goes without saying that I have almost from the first use set LMU very well in all of these areas, whether it is FFB, graphics, sound, file, etc.
And I have actually improved these settings each time I start LMU.

Which simply means that I have used LMU truly in the best conditions thanks to my extensive knowledge of rFactor 2, unlike probably at least half of LMU users. So I have used it in very good conditions.

And there again as for what I said for realism and physics, there are as I said many small improvements here and there in all areas but still maybe not all areas but almost, but I absolutely do not agree that these improvements are significant.

For example, for the graphics, there are a whole series of small, appreciable things that are improved, but nothing extraordinary, the result is a better experience but nothing more.

For the sound, it's exactly the same, with the same sound engine as rFactor 2, with the difference that the homogenization and exploitation of the sound engine by the entire content of LMU unlike rFactor 2 plays a major role.

As I explained, the UI has exactly the same parameters with a few small exceptions, the adjustment possibilities are almost identical, the UI technology is quite probably identical I presume, but imagine that I find it much less good than that of rFactor 2 for example !

The level of detail of the circuit is a little better but nothing impressive, it's still the same graphics engine, the track scan looks approximately identical in terms of physical precision, the graphics of the cars are at the highest level of rFactor 2.

But all this is not of much use to me since it is the only game that could have interested me outside of rFactor 2, the physics do not impress me at all since it is almost identical or identical, and concerning the content I really find it extremely uninteresting :

Ultra recent cars with driving aids, a lot of aerodynamics, very few different styles and types of cars, cars from the same years and very clearly not from the most interesting years in terms of driving, that's the least we can say, really the least we can say...

So of course LMU is not at all a bad game in reality, but the added value of playing it for me extremely low if not zero (because rFactor 2 exist).

And indeed, there are indeed less factual elements of the order of affect which play a role for me, namely that the arrival of LMU has quite simply killed rfactor 2 on a large number of aspects, and I am not satisfied with that, sorry.

 

[AMillward]

Except that I think that many people are far too influenced by what the developers say, these are great new tires blah blah blah, I do not believe for a single second, for example, that the tire model has really progressed significantly.
Just see how the initial tire model of rFactor 2 has progressed extremely slowly over the years, I honestly do not believe it, and I think there is a placebo effect of sorts.

you have to get some heat in the tyres on LMU before you can start setting some fast lap times, something you never don’t really have to worry about with rf2. If anything the tyres have been made more realistic.

the tyres HAVE changed. It’s not a placebo at all. If you try to lob a GTE into turn one of Portimao on the cold tyres, it’s going to struggle. After a couple laps it starts to feel a lot better.

 

[Lazza]

@pilAUTO for anyone to say LMU is leaps and bounds ahead of rF2 - especially regarding physics - would suggest rF2 isn't very good.

The differences are subtle because they have to be.

And frankly the "LMU killed rF2" line is getting very old. rF2 had 12 years to become a success, and isn't actually any more dead now than rF1 was in late 2007 (when updates stopped, and modders took advantage of the stable base to make the best content for years).

You like rF2, that's great. It doesn't mean everything else has to be bad just because it's not rF2. That even includes sims from other companies *gasp*

 

[AMillward]

At the risk of sounding mean, if people think that the tyres are exactly the same and physics in LMU are exactly the same as rf2 they either:

1. Haven’t actually played LMU
2. Have only spent 5mins in it
3. Are saying this because they are trying to justify their “rf2 best” biases.
4. Aren’t driving fast enough notice that the tyres in LMU need to warm up.

You like rF2, that's great. It doesn't mean everything else has to be bad just because it's not rF2. That even includes sims from other companies *gasp*

This.

 

[8Ball]

No different to GTR2 vs rFactor sim wars ( shrugs )

Only change now it's S397 not ISI aka LMU vs rFactor 2

They all use ISIMotor.....oh wait

 

[avenger82]

If it were that easy, they'd do it. Heck, if it were that easy we'd all be making sims.

Not necessarily , as I heard/read a few times S397 said something is not getting fixed or implemented because it’s hard. Then you see e.g @redapg fixes it in no time. IIRC that was the case with some bug in H pattern shifter in Caterham

 

[hitm4k3r]

Not necessarily , as I heard/read a few times S397 said something is not getting fixed or implemented because it’s hard. Then you see e.g @redapg fixes it in no time. IIRC that was the case with some bug in H pattern shifter in Caterham

I hope I don't sound disrespectfull for saying that but it happens with pretty much all moddable titles at one point or another that certain things that were deemed difficult by the developers get implemented or improved by people in the community. One explaination for this is that modders usualy don't have to work with budgets, time constraints, in other areas of the software or with the element of risk. You simply have to pay someone to do the fix, while there are other tasks aswell.

I modded Battlefield 2 for a long time and still do once in a while. The stuff that we do today has nothing to do with what DICE intended back in the day. A big part of that is due to our lack of time constraints and the freedom to experiment. You rarely have that comfort when working for a software house. I think this often get's lost in all these discussions but the people working on those sim racing titles - no matter if it's ISI, S397, Kunos, Reiza, Sector3/KW Studios, Marcel with his new product or even the people from Rennsport - all those people are very intelligent people and not too dumb to do certain things. They simply have to balance costs versus ROI alot of the time.

 

[redapg]

Not necessarily , as I heard/read a few times S397 said something is not getting fixed or implemented because it’s hard. Then you see e.g @redapg fixes it in no time. IIRC that was the case with some bug in H pattern shifter in Caterham

I think i must say something here.
I didn't fix it, i just completely deactivated the Gearprotection of the Mini and the Caterham.
To rewrite the Game Code, to really fix it would be much more complicated, i guess.
But maybe we can be happy, that it wasn't fixed by Studio S397, because how often have we seen, that X Things were fixed, but X other Things got broken at the same Time.

 

[avenger82]

Ok I stand corrected. I only remember reading it on this forum that people were happy with the “fix” anyway.
Another fix that we heard is too hard - otherwise it would be fixed long time ago is the wheel self centering on race restart in single player. It’s a big problem with strong DD wheels and I guess it will never be fixed/ameliorated despite somehow in multiplayer (e.g. on RaceControl servers) it doesn’t happen .
RaceRoom had similar problem and it took a long time , but they finally fixed it. I think AMS1 had similar issue too so it’s probably old problem dating back to rF1. Some simple ideas/workarounds were proposed in the long thread , but it’s still on the TODO list.
If someday rF2 code would be open sourced (very unlikely) then such issues would be fixed by the community.

 

[Bernat]

Sometimes the fixes aren't as easy as they might seem from outside. Some want the wheel to autocenter automatically, some don't, then in some cases it will autocenter while in some others it won't, maybe related to system drivers or the hardware. There's no line of code saying "if user/wheel is ... then break the game", it's far more complex than that.

Of course with lots of time, skill and knowledge, most bugs can be fixed, but the fix isn't always trivial. Some users might say it's more broken than it was while some might say it's a fix.

 

[Lazza]

Ok I stand corrected. I only remember reading it on this forum that people were happy with the “fix” anyway.
Another fix that we heard is too hard - otherwise it would be fixed long time ago is the wheel self centering on race restart in single player. It’s a big problem with strong DD wheels and I guess it will never be fixed/ameliorated despite somehow in multiplayer (e.g. on RaceControl servers) it doesn’t happen .
RaceRoom had similar problem and it took a long time , but they finally fixed it. I think AMS1 had similar issue too so it’s probably old problem dating back to rF1. Some simple ideas/workarounds were proposed in the long thread , but it’s still on the TODO list.
If someday rF2 code would be open sourced (very unlikely) then such issues would be fixed by the community.

I'll be honest, there are lots of small things that would very likely not take that long to do but have never happened. And while we can't see the code, we can see enough of the implementation (parameters etc) to judge what the code has access to and could utilize. The caterham/mini shifting bug (sequential only) looks like a logic issue that would merely take someone looking at for a while, but for whatever reason it's never been a priority. (and actually that's not right at the "easy" end of things I've seen over the years, but for sure it's got to be simpler than tuning AI behaviour for example).

So: it not happening doesn't mean it's necessarily too hard to do, but being users we just don't get what we want sometimes.

When I speak of LMU improvements I'm taking into account the nature of the changes, the context of a new title, and comments I've seen (public and private) from devs. Bringing some of those changes back could easily present a heap of challenges, potential new bugs, and for what reward? Happier users, sure, but probably not many new buyers of what is now a cheap game. The return just isn't there.

 

[avenger82]

Some want the wheel to autocenter automatically, some don't, then in some cases it will autocenter while in some others it won't

Well that seems easy, just add an option: Do not EVER autocenter the wheel. I think this solution was mentioned in the long thread regarding this issue.
I get that perhaps some functions might not be called directly, but rather indirectly via Direct input API etc. , but somehow other games don’t have this issue using the same API.

 

[avenger82]

My wild guess is that vast majority of the code core ,( excluding GUI etc. ) is still the same in LMU. You can’t rewrite a game engine, AI, netcode etc. that was developed for over 20 years (including rF1 development) in one year. Granted that after S397 took over, development of gMotor engine was slow and AFAIK even last couple years at ISI it was almost in a maintenance mode.

 

[8Ball]

I agree in principle however ISI were first to introduce a number of features that some studios still don't have.
Every other engine I drive still has robots for AI ( shrugs )

So imho they are the ones need to catch up to faulty 20 year old engine, not the other way around.

iRacing Madness Unreal whatever.. they all have glitches, bugs, missing features things they do well things they do bad and some things they don't do at all.

Peace

 

[hitm4k3r]

My wild guess is that vast majority of the code core ,( excluding GUI etc. ) is still the same in LMU. You can’t rewrite a game engine, AI, netcode etc. that was developed for over 20 years (including rF1 development) in one year. Granted that after S397 took over, development of gMotor engine was slow and AFAIK even last couple years at ISI it was almost in a maintenance mode.

I sometimes read statements like yours and I wonder how people meassure these things. What exactly means "development of gMotor engine was slow"? For the fact that we got a rewritten graphics engine with a new rendering pipeline, a new sound engine, new netcode for LMU, new webbased UI infrastructure, a complete online ranking system with matchmaking, Real Road 2.0, several rewrites of the AI logic - I think this engine made huge steps during the time that S397 had it in their hands and considering the size of the team working on it. This would also imply that you know how fast the other devs are working, wich is basicly impossible. I can only give you the advice to boot the last ISI build from 2015/2016 and see where LMU is now.

 

[avenger82]

I mostly meant core gMotor engine e.g. physics (netcode etc.) development was slow. For me that’s the most important aspect of a sim and that’s why forgot to say that yes, graphics was significantly improved by S397. At first the move to DX11 by itself didn’t wasn’t revolutionary , but over time incremental upgrades made tracks like Nords look modern, even one of the best in sims.
UI finally looks modern (especially in comparison to previous 90s UI) and mostly ergonomic (much better overall than first iterations of HTML based UI). It’s OK, but still often sluggish in rF2 and not that smooth like in ACC or pCars2 etc.
What do you mean we have new netcode and sound engine for LMU? Was it rewritten? I know netcode was improved. Still not enough to host Virtual LM , though (yeah I know that second time, officially there was DOS attack fault )
Also several rewrites of AI logic? Do you have the source for that? Anyway, IMHO AI was a little better before S397 took over . They fixed something, but broke something else.
UI and ranked multiplayer and I think AI is not part of gMotor engine.