Gsce bonanza

It's literally rFactor 1 physics. I can easily tell especially once you start pushing the cars and playing with the limits. Niels has worked magic, but it still behaves like pure RF1 and that will not change no matter the car/mod, no matter the genius work and numbers Niels "plugs into" the engine. The way you rotate cars around corners, the way cars behave during oversteer, the way grip is lost and regained, the way oversteer is corrected, the way front tyres grip/slip at lower speeds, etc. etc. etc. it's still just regular 'ol ISI/RF1. Niels has "massaged" it absolutely amazingly but don't be decieved, it's still the same 'ol. The nice FFB helps further to decieve but he vehicle behaviour is still pure ISI/RF1.

I can't see Niels sticking with just RF1 physics code now that Reiza have the rights to modify the core coding/engine/EXE itself. I bet there will be core physics engine (including tyre model) changes and improvements (and I don't mean "features" like track-rubbering and flat-spots but actually physics).

I have a feeling most of the core physics changes will arrive for their next game in 2017 but hopefully they can do some core physics changes for SCE/AM too and then people can't say it's just RF1 anymore.

 

Are you the same Spinelli from RD and ISI forums?The one wich posted that superb guide to tweak graphic on isi based games?The one wich make me smile every time I start rf, rf2 and gsce?Are you?

Maybe is placebo (or I am a total noob) but I found sce cars more grippy than rf ones?I can brake and turn more easily, avoiding to spin more easily than rf...

 

Cars in SCE feels like there's more grip before the limit. When you step over it, SCE begins to feel like an rF1.

 

GCSE soon to be Automobilista and AC combined give me exactly what I want from my simracing, nothing else left to say

 

Some fuel into the speculation fire: Reiza state pretty clearly on the Atomobilista page that there are improvements in the pipeline that will be released with Automobilista and NOT be included in SCE.

So if I was them I wouldn't bother improving the old title and put all the energy in the new one UNLESS I could transfer the software modules, findings etc. into the new one with minimum effort. I think Automobilista will sort of be their playground for the new title and that there will be a significant physics engine overlap.

 

Yeah, judging by the timelines alone would support that theory. What's not clear to me is what rF engine are they actually using now, are they updating to rF2, certain aspects of it or just adding to rF1 using the base code?

 

The way it reads to me they are not really making any physics changes the tire model. Load sens, camber gain, combination of lat and long grip etc stay as they are. They are adding more stuff to it. I think tires getting dirty was mentioned at one point (or is it already in there) and debris and marbles. While I think rf physics engine is very good even today it has some issues which I personally find really annoying at times and I'd hope reiza would fix or improve some of those.

Also being a game from 2005 I'd imagine they could turn on or turn up some settings in the physics files for example. Like sampling rates for the physics. Ac for example runs at 333Hz I think. No idea about rfactor but surely there should be room to improve. LFS does or did run its tires at 2000Hz for example and other parts at lower rates. Of course it is not a simple thing of changing one line in the code and expecting things to just improve without issues but I'd hope reiza not just adds features but develops the base tire physics stuff too.

As for my major issue with rfactor I'm starting to believe it is down the way the physics engine and graphical engine are connected somehow. Like when driving the graphical engine doesn't show the full roll of the car in corners for example which makes it insanely difficult (for me) to figure out when the car is coming out from slide. Over the top of my head I'd imagine the graphical engine doesn't show tire deflection at all. I'm not talking about tires visually deflecting like in rf2 or lfs but simply in corners the view rolling (when driving) sideways little bit more because not only the suspension is compressing but the tires too. For me there has always been something missing in all isi based sims and I'm not the only one.

Rf2 is little bit better in that regard but nowhere close to ac or iracing where I can feel the tires starting to grip really well (although in iracing I'd still probably spin ).

I hope reiza does it smart though. With their small crew they probably want to avoid making physics changes many times because they need to update their current cars as well.

 

Anybody knows more precise than Q1 2016 release time for AMS? Will it be Early Access or full fledged title?

 

Full fledged title. 30-60 days from the announcement. One of the dev said it on racedepartment.

 

Half a year is not much for in depth changes to the engine. They shown a few images where you can see how much work they did with the graphics. Also this will be only a "tempolary" title before the new reiza17 game. So I think they just want to do the most important stuffs so we as players can test it out. And these improvements will be the base for their next game. The real and more in depth changes will be in the reiza17 game IF there will be any... They said that the reiza17 game will have DX11 or 12 graphics don't remember. So I assume they want to do dramatic changes in the engine.

Also this 10years old engine and 2005 game talk is really getting old. Unreal engine 3 is also a 10 years old engine and still going strong. ISI Gmotor 2.0 is still a good engine. And now a dev team want and can improve it.

 

AMS is using isi motor 2.0 still. The rFactor 1 engine. However, over the summer Reiza bought the source code to the isi motor 2.0 so they can now change or add features at a fundamental level, instead of tacking them on via code injection like they had done for all of the lifespan of GSCE so far. Simply put, this will allow them to evolve the isi motor 2.0, much like ISI has done with isi motor 2.5 which is the rF2 engine. Reiza will be adding tire damage and flex like rF2 has, but in their own way. Other features will similarly be added, so to say this engine is 10 years old is a bit ignorant now without the context of what is actually happening. (I dont mean ignorant to be offensive, just that with extra knowledge, they arent the same thing)

Reiza has stated that the crowdfunding backers will get the beta version of AMS at the end of january. In the past, backers have had beta access for roughly a month before the full public release. So, I would expect AMS to launch officially at the end of February/early March. Near the end of Q1.

Well, as I outlined above, Reiza has the source code for the game engine they are using now. And can evolve that engine as they need to by adding features as they need to.

AMS will just be GSCE, renamed, and with all of the extra content from the crowd funding campaign. Along with a few new features like the dynamic road they previewed, tire flex, and other stuff.

Their next game, tentatively scheduled for a Q1 2017 launch, will of course just be a further evolution of GSCE/AMS and will feature even further improvements like dynamic weather and a DX12 graphics renderer. The game will most likely feel quite similar to GSCE/AMS since it will still be isi motor 2.0 (as far as I can tell), just heavily improved since they can easily do what they need to with the source code to improve their future endeavors. Again, much like ISI has done with the transition from rF1/isi motor 2.0 and rF2/isi motor 2.5.

Hopefully this helps clear some things up.

Cheers

Edit:

You could say the same thing about rF2 and isi motor 2.5 then. Since that game is also based on an engine from 10 years ago. The difference however, is that like ISI, Reiza has source code access so they can now add new features and make improvements that are light years away from the original release of rF1 10 yrs ago.

It is just an evolution of the same engine, with massive improvements. Although, Im not comparing GSCE/AMS to rF2, since ISI obviously has done a lot of work from their rF1 days. Reiza is now beginning the same process.

 

well it kinda makes sense,
why would you want the loose surface physics in SCE when there are no dirt tracks in there ?

 

Haha, yes, that is most likely me. Thanks for the kind words

+1
Niels/Reiza's fantastic work make the cars wonderful to drive and feel solid and grippy (the FFB "feel" also gives this impression thanks to good FFB implementation by Niels/Reiza). Niels really is "the man" in my opinion. However, as you say, once you start battling and playing with the limits, getting slip, "leaning" on the tyres' limits, etc. then the vehicle behaviour and actions and reactions are just the same 'ol RF1 which is extremely frustrating and unnatural/"digital" in my opinion. You can sense this regardless of if it's the modern street-car Camaro, the 70s touring car, a modern GT racer, or 90s F1 car.

Once Niels/Reiza start updating the physics engine, I expect massive improvements to the vehicle dynamics and kinematics as, in my opinion, there are some very odd/wrong moments in the ISI physics engine that have been there all the way back to even F1 2002, and still there 15 years later in rFactor 2.

No sim is perfect, obviously; not AC, NKP, LFS, IR, SCE/AM, RF1, RF2, RBR, PCars, etc. That's not the frustrating thing though as I, nor should anyone, expect perfection. What is frustrating though (and extremely frustrating at that), is when at-least 15 years of evolution goes by (F1 2002 to rFactor 2) and the same oddities in vehicle behaviour are still present despite 15 years of core physics engine (including tyre model) evolution. It actually blows my mind.

 

maybe if you were actual developer, you wouldn't be so surprised
15 years that's like 5 games, or 3 if it takes longer, so let's say 4 (average)
you create 1 engine from scratch
second game you try improve but mostly re-use ( want to make extra money and offset some cost of the 1 engine)
third game you decide to start form scratch , again huge costs , and not polished version
4th game you are improving and polishing previous generation,

so yes, no point of expecting perfection , plus as you say it's a lot about "feeling" , and perhaps devs feel that the cars behave the way they should , which is why they won't change some of the characteristis that some other consider wrong etc

 

What I meant was that the rf engine reiza uses was literally 10 years old code. It's source code was last time edited some 10 years ago by isi and not been edited ever since (much) until reiza bought it. Sure in the meanwhile simbin/blimey/slightlymad/raceroom has used the same code but reiza did not buy the code from any of those. It is not about how long it has been around (it is probably closer to 20 years old). I think reiza did ask isi to do some changes to the code along the way but in the end it is literally 10 years old code for 99.5% part. It is not a bad engine but it is old.

To put it into perspective the code reiza bought is older than nkpro at this point. During the time from the last patch release of rfactor and today kunos has upgraded the nkpro engine and worked on it more than isi has worked on the rf1 engine. It is really just a fact, nothing to get upset about especially when most people don't even mean anything negative when they say it. And of course now when reiza has its hands on it things should ... let's use the word ... happen.

-

When it comes to developers it is not that uncommon for the end-user to think that something is broken whereas the developer sees it as working as intended. It is not an error - it is a feature - kind of thing. But even then I really hope the issues with rf1/rf2 engines is not about how some developer thinks it should feel. I really hope it is something in the code that can be improved once people figure out what it is (it probably is not just one thing).

 

I'm sure there are some SW guys in here that will understand that using a solid baseline of old and proven source code, functions, modules or even just the flow charts of those is not a bad thing, at all - quite the opposite is the case. "All Reiza has to do" is to add the functionality and accuracy that could not technically be achieved due to PC ressources back in the day. If the code is well done and structured (I assume it is given what Reiza did even just by code injection so far) this is no problem, at all and you can literally keep doing this for decades.

Now that was my final wisecracking outburst on this topic ;-)

 

You can't keep using old engine for eternity without having to occasionally do big rewrites of some parts of the code. Rf for example does not support 64bit, it can only do dx9 and it can't support multithreaded cpus. Being an old engine does have drawbacks no matter how good the engine is otherwise.

 

I don't think ISI has ever started over from scratch. I'm pretty sure F1 2002 (and maybe prior with the late 90s SCGT) all the way to the current build of rFactor 2 has been continual evolution of the 1990s engine except, maybe, the tyre model in RF2 which seems all new (even though there are still many similarities of physics from RF2 all the way back to F1 2002 which leads me to believe that either the end-result of the new tyre model isn't as new as the tyre model itself, or it has nothing to do with the tyre model but with the remaining aspects of the physics engine).

 

Starting from scratch in software development seems to be very rare. I'm sure AC has plenty of NK Pro hidden inside it.

 

Here is a 10 years old code:

public class Hello1
{
public static void Main()
{
System.Console.WriteLine("Hello, World!");
}
}

Still doing fine.

I'm sure reiza competend enought to judge where the code needs to be changed or extended. 10 years old engine still doing fine. DX9 can be really nice. The thing is: The nice game graphics is unrealistic because of the unrealistic posteffect image quality standards.

They bought it, they will update it and that's it people overthink it.