Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ZeosPantera]

This is solely about a simulator product that would be using the engine, and given its specific properties it could combine several simulators into one

I'm basically asking what's the market potential here.

Like, is the sim world comprised of disjoint groups of hardcore sim fans of their specific genres, so that the product would be getting equal bashing from all sides because not being focused on their sim domain

Truthfully, once you have the world and as accurate a physics model as you can insert into that world all other elements for all sim types should fall in place. Gravity for a car is the same as gravity for a plane is the same as gravity for a submarine and if the physics are correct gravity affecting an orbiting space-craft should be the same gravity just less of it due to distance.

I never really understood how "sims" could all be so different. The forces around us are pretty constant. If 9.8 m/s^2 is the force then it should be the force in every sim. I can understand how not having severe collision modeling in a pure flight sim would be simpler because noone is suppose to crash, but why not work on a complete physics engine in the first place.

Driving sims have to either have incredible collision instance tracking so that 4 spinning tires of a certain compound interacting with a surface material cause a car to move forward and turn(RBR). Some older and cheaper games just fake it all with vector data and throttle positioning (GTA4). But a driving sim would also need to calculate wind resistance and downforce from a spoiler or splitter so wouldn't that mean it needs to do all the calculations a flight sim currently does to reproduce that aspect correctly?

I say all or nothing. No point in simulating the whole world and limiting it to only one aspect of anything. You should literally sell the whole empty world to us and we can just install all the golf, driving, sea exploration, helicopter racing, mountain climbing, bobsledding, baseball, railroad, cattle-driving, curling, murdering sim modules we want to use the world and its physics.

I imagine something along the lines of "Steam" You download steam and you get access to a huge list of games. Why not subscribe and download Outerra, Get the world and all its updates and then just buy modules that can be added. A separate company could then put their module with all the models and control configs up for sale. You buy and add it to your outerra world. Even a simple puzzle game could just exist in the middle of the desert somewhere. Right next to my race-track and WWII tanks.

Multi-player is however what outerra should really focus on. No point in simulating the earth and 2.2 billion km into space if it will be you alone. I have to think 50,000 people roaming around in the outerra world would still leave plenty of elbow room. Just break it into different dimensions if need be. I really am talking about world changing possibilities here. Or maybe I am just rambling.

[cameni]

Well, a simulation is just an approximation of reality, a complete physics engine would be infinitely slow to both make and run
But I'm sure you didn't mean that completeness

Still, it's always a compromise. RoR guys are simulating the vehicles with advanced beam physics that can make a realistic simulation and deformation (with proper models). But it is also quite demanding, able to down any system easily, to the point it ceases to be fun.
Except for hardcore fans who favor sampled realism over the fun, that is.

But I say it doesn't matter. We want to create a platform where even such physics could be run. We want to create a world that is needed by every one such simulation or by the games that want to play on whole planets with detail at all levels. And let's make it open in a way allowing to plug-in any simulation system, that would be lightened from the burden of having to create a believable world every time.

However, I think the platform should come with a good implementation for basic vehicle classes to successfully jumpstart.

Other points are valid - the platform should be available for both developers and users, and it should provide the market place for addons etc.
And the multiplayer, yes it will be very important too.

The different dimensions should be the separate realms where one can be, the overlays defining in what epoch the world is and how it is changed from its natural state (=civilization). Crossing between the realms will be limited, at least for the active roles (unless it's a part of a game), but for the passive viewers it might be quite a fun to cross over to the age of dinosaurs with a jet fighter

[ZeosPantera]

the age of dinosaurs with a jet fighter

You had me at dinosaurs...

[corona]

Wow, cameni, don't get me wrong, the epoch thing sounds super duper awesome, and I really mean that.

But don't aim too high at first. It worries me that it'll be another DNF (wow I just release dnf = did not finish and duke nukem forever.....rofl). If you haven't done so already, clearly define, right now, exactly what will be in, lets call it Engine V0.1 or whatever that you release first.

And I fully agree with your other points, provide a baseline physics model, but keep everything open.

[cameni]

Don't be afraid, we have enough professional experience to know not to aim too high Erm .. not too often?
But also from the same experience I got used to outlining the path to the end goals in advance, just to have it on mind when designing the elements of the project.

For example here it means that when we'll be doing a persistence layer, for storing and loading the world data on servers, we'll be making it with the realm support in mind, even when initially there will be just a single world. Otherwise we'd have to redesign it later for sure. So I want to include it into the code paths that are clear and independent now, and to keep it in mind when designing those not so clear.

Because it's extremely easy to simplify things in a way that later really complicates everything. On the other hand, it's always about finding a good balance between what one can dream about and what he can do when awake

[tasmanet]

Hi Guys

I have just landed here after giving Mathijs Kok a bit of a stir over at Aerosoft for "Stuffing People Around" .Personally I dont think Aerosoft will ever get there.

My only effort was creating "WDW2012" for FS2004 allowing for only 5 EMails must have been half decent or extremely bad.
But  AVSIM has lost the files and I dont have time to upload a newer version
So I have built a FERRY and Monorails from scratch , designed a new A380 Cockpit, modified FX's,made AI plus of course all the buildings/scenery etc around the Magic Kindom Area.

I get what you are stating about an all inclusive FS World with sub contractors / licensees developing their own Worlds within  same FS World

http://www.forum.aerosoft.com/index.php?showtopic=30062&st=328

Got to go and will ponder about some of the above posts

Tas

[cameni]

Yeah I see the forums are getting pretty cross-linked back and forth
But apparently mainly the people with inclination towards multi-sims are coming here, which I consider a good thing.

[Michal]

I'm interested in two or three types of simulated vehicles. Aircrafts (including helicopters and lighter than air airships), trains and ships. FSX with high quality aircrafts is my primary simulator (I have few additional aircrafts - most of them have high-end avionics). I like to play Railworks and thinking about getting Ship Simulator too. Two of these three games have one drawback - they don't have whole world to play.

I was waiting for Microsofts Train Simulator 2 (the second one, based on FSX engine). They had really awesome idea behind this title:

Q: What is the "World of Rails" feature?
A: High Accuracy Route = track accuracy and geospecific scenery. World of Rails = corridor accuracy and geotypical scenery. What the World of Rails means for players is that you will be able to drive a train anywhere in our default set of rails around the world and can keep driving for as far as the track extends without having to leave the game to load a new level. For builders, it means you have a head-start of good-quality digital elevation models, landclass, and demographic data from which to start building a route, but you can also start from scratch or build a fantasy route.

http://tsinsider.com/



This is how I'd see multisimulator - as an engine to play and build for community. It would require scripting engine for additional vehicle systems (like avionics in aircrafts, or complex engine management for steam locomotives etc.). Maybe it would be possible to expand script interpreter to scenery too - like ground vehicles on airports, ATC, or signalling and dispatcher for rail simulator. I don't know if it's possible, but it would be nice to have

[cameni]

I've been always wondering how the FSX engine would fare in a train simulator, given it would be all about ground level detail.

This is how I'd see multisimulator - as an engine to play and build for community. It would require scripting engine for additional vehicle systems (like avionics in aircrafts, or complex engine management for steam locomotives etc.). Maybe it would be possible to expand script interpreter to scenery too - like ground vehicles on airports, ATC, or signalling and dispatcher for rail simulator. I don't know if it's possible, but it would be nice to have

It doesn't necessarily have to be a scripting language to handle this, rather the plugin architecture should be designed to handle it (it may delegate it to a script, of course, but the core functionality should be compiled for performance reasons). I'd guess that a creator of an airport will code the vehicles and signals etc as well.

[MatthewS]

I've been always wondering how the FSX engine would fare in a train simulator, given it would be all about ground level detail.

ACES had done a lot of work on enhancing the ground level experience.... FSX and TS2 were based on the same "core" engine, with extra stuff added in to the core engine to cater for TS2 (eg animated people on stations/trains etc).  Apparently they had a core engine team as well as the TS2 team.  Then FS11 would have resulted in the core engine being enhanced again.

Gee TS2 would have already been released over 18months ago if Steve Balmer had not pulled the plug and ACES would have only been 6 months from releasing FS11.  

In another universe some lucky buggers are TS2 simming right now!  

It doesn't necessarily have to be a scripting language to handle this, rather the plugin architecture should be designed to handle it (it may delegate it to a script, of course, but the core functionality should be compiled for performance reasons). I'd guess that a creator of an airport will code the vehicles and signals etc as well.

Yes IMHO just have a flexible plugin API and then developers can use what they want, C++, C#, Perl etc.  Integrated scripting at the expense of a plugin API is never going to be attractive to commercial developers.

[cameni]

ACES had done a lot of work on enhancing the ground level experience.... FSX and TS2 were based on the same "core" engine, with extra stuff added in to the core engine to cater for TS2 (eg animated people on stations/trains etc).  Apparently they had a core engine team as well as the TS2 team.  Then FS11 would have resulted in the core engine being enhanced again.

But that's mainly for modeled areas like stations or the track corridors, isn't it? If you created a custom track elsewhere you'd have to model everything again. And it needs a huge amount of data to be modeled to enjoy and believe the world.

 I perceive it as a big advantage of a procedurally pre-modeled world that the terra incognita will be already looking natural, and you only "civilize" it. Not every one will like some implications of this approach, though.

[Michal]

Whole world was supposed to be populated with generic data - landclass (forests, meadows, urban area) and vector - tracks, roads, rivers, lakes and seashores. In Trainz or RailWorks you have empty flat area. Route developer has to place every single tree, every meter of track, every road or river. There is even no height data. World of Rails was a huge step in right direction. I think it would be possible to place a railstations, signalling, and you have believable prototype of nice route. This approach is impossible in RailWorks or Trainz. That's why I think Outerra has a huge potential. In fact your flat trees already looks better than forests in FSX, Trainz, or RailWorks. If you'll be able to make a core platform for open simulator (open means open architecture, not open source), I'm sold.

About scripting engine - scripting (like xml in FSX or Railworks) is much easier to learn for non-programmers like myself. I can make complete avionics for my planes in xml, but I have no idea how to make the same in C++ I believe this is main reason why FSX developers are ignoring XPlane. Because of it's plugin system.

[cameni]

Whole world was supposed to be populated with generic data - landclass (forests, meadows, urban area) and vector - tracks, roads, rivers, lakes and seashores.

Ah, I see. But wouldn't the default resolution of landclass and elevation data be unsatisfactory for a train sim? I mean, for a flight sim it is already nothing much when low flying, without some large addons providing more detail for a specific area. I can't imagine how the generic terrain would look in a train simulator then.

About scripting engine - scripting (like xml in FSX or Railworks) is much easier to learn for non-programmers like myself. I can make complete avionics for my planes in xml, but I have no idea how to make the same in C++ I believe this is main reason why FSX developers are ignoring XPlane. Because of it's plugin system.

I understand; what I wanted to say is that the core system won't be directly interacting with scripting, rather that the ability to script something will be tied to a particular plugin or interface itself. Scripting is less generic, easier because it's designed for a particular purpose, so there's no point making it overly generic within the architecture. However, a unified scripting subsystem should be available to all plugin makers so they can expose scriptable parts in a universal way. Like, scripting the avionics or the behavior of ground vehicles on an airport, and such.

[Michal]

Here are few screenshots without custom objects, only World of Rails data:

http://tsinsider.com/Articles/AboutTS2/WorldOfRails.aspx

It looks good in my opinion. I've read on train sim forums, that Aces were working on new lighting system, not implemented on these screenshots yet.

I think it should be possible to make something like this with Open Street Map. There are raw vector data for rail tracks. Engine should be able to automatically add junctions, crossings, etc. It should be possible to change tracks layout by hand in some kind of editor.

[cameni]

Not bad, although those screens can hide quite a lot. Popping objects and trees, for example.
Anyhow, it doesn't matter now :/

Yes, OSM should be usable here.