Winterleaf Notes on OMNI Engine (Formerly DNT)
by Vince Gee · 10/15/2013 (5:37 am) · 39 comments
Before I go into detail about the answers to your questions I want to point out a couple things.
Currently, there are two release versions of DNT. DNT 1.0 and 1.1.
DNT 1.0 is built on top of T3D MIT 1.2
DNT 1.1 is built on top of T3D MIT 1.3.
These builds do perform better than Stock T3D but the CSharp scripting resembles TorqueScript and tend to be a tad wordy.
These versions of DNT are more like T3D with CSharp scripting, they aren`t very object oriented.
Also, these versions are not MONO compatible.
We do have a couple game companies building games using these versions, but the technology is about 8-10 months old.
OMNI (Which is the next version of DNT), which will be released Q1 is a rewrite of DNT as well as many parts of the T3D engine.
It has among other enhancements,
It isn`t an easy endeavor, especially if it is your first attempt. If you use a lot of namespaces, super-classes, etc. it can be quite tricky to translate it to CSharp. There are features which simplify the process but a solid understanding of CSharp (and inheritance and object casting) is a requirement. I have completely converted the Starter example at least 4 or 5 times. The last time I converted it, it took about a week. The first time I converted it, it probably took me 2 months. (On the plus side, you do get priority email and forum support with your purchase and we usually get back to you within 24 hours or less). And usually, if I can`t resolve it over the Forums or Email, I will remote to your development machine and look over your code. Plus I plan to have a SDK with examples on release.
The main thing that has really spiked my interest is the cross-platform support
Yes, our goal is to have a platform where you can build your game on MS Windows and then just recompile it on a supported platform and you will be able to release your game with little extra work. Currently, everything we write is MONO/Linux compatible and we do have a full Client/Server version working of DNT(OMNI) running on Linux. We are working very hard at pushing out an Android and MS Surface build.
In regards to performance, OMNI performs reliably faster than stock T3D.
Under no load running an empty mission we see about a solid 21% increase in FPS and 16.21% increase in MSPF over stock.
Where OMNI really shines though is under load.
One of our load tests is bring up a zone with 300 AI in it. The AI are split into 2 or more teams and constantly attack each other. A dedicated server (Core 2.3ghz) runs this zone at about 40% CPU utilization. The dedicated clients run upward of 60+ fps.
When running the same load test as a stand-alone we can usually run about 100 AI in the zone at 30-80 FPS.
Our test machine is an old Quad Core 2.3 Ghz dell computer, average AMD video card with 1 gig of ram.
Running with all Shaders turned on is also day and night between OMNI and stock T3D in regards to frame rate and MSPF.
OMNI can run TorqueScript, but yes, in general the resources need to be converted to CSharp.
We are working on incorporating the open source resources like GMK, Path Finding, etc. at launch, 3rd Party add-ons will take conversion.
In regards to patching of MIT fixes, we have a full time programmer who merges in the changes from the committee into our build. So at the very most we are a week or so behind the MIT build. Our code base at places is significantly different than stock MIT, but we try to evaluate each MIT patch for it`s worth and application. Some things done in the MIT branch just don`t make since, or are implemented poorly/incomplete. We fix these things by ensuring that networking/tools/etc. are upgraded to when the MIT patch misses one of them.
We couldn`t agree more.
One of the things that always killed me with stock T3D was that TorqueScript is proprietary. I cannot use any code I write in TorqueScript for anything else other than T3D. I have written tons of CSharp classes and code snippets over the years and it`s nice to know that I can just cut and paste them into the code w/out worrying about converting the syntax.
One other advantage we have noticed is that programmers working on our games are more apt to learn CSharp than TorqueScript. The CSharp language is a tool that can be used outside of the game development world, So learning it isn`t considered wasted time.
One final note on this is also the plethora of CSharp code available on the Internet. You can pretty much find code to do anything written in CSharp. It also simplifies things like database integration and 3rd party drivers (For use in things like robotics). If there is a CSharp driver, you`re in business. It opens up a whole new world to T3D, one where the limitations of engine integration no longer matter.
Vince Gee
Winterleaf Entertainment L.L.C.
Currently, there are two release versions of DNT. DNT 1.0 and 1.1.
DNT 1.0 is built on top of T3D MIT 1.2
DNT 1.1 is built on top of T3D MIT 1.3.
These builds do perform better than Stock T3D but the CSharp scripting resembles TorqueScript and tend to be a tad wordy.
These versions of DNT are more like T3D with CSharp scripting, they aren`t very object oriented.
Also, these versions are not MONO compatible.
We do have a couple game companies building games using these versions, but the technology is about 8-10 months old.
OMNI (Which is the next version of DNT), which will be released Q1 is a rewrite of DNT as well as many parts of the T3D engine.
It has among other enhancements,
Complete CSharp Integration
Syntactical sugar
- Implicit/explicit casting
- Helper functions to simplify object creation etc.
- String extension to reduce CSharp boxing.
All SimObjects have:
- Member variables
- Events
- Inheritance
- Caching
- Type Converters
All Script enumerations are exported
Improved inline callback system
- No longer bouncing through the console
- General optimizations for performance.
Thread safe implementation by locking the `Tick`
String Extensions
- Convert any SimObject to string and vice-versa
- Convert any SimObject to int and vice-versa
- Convert any SimObject to uint and vice-versa
- Convert any base type to T3D compatible type and vice-versa
- Complete C++ Code Parser and CSharp code generator (DNTC)
Reads and parses
- InitPersists
- ConsoleFunctions
- ConsoleMethod
- DefineEngineFunction
- DefineConsoleFunction
- Implement_Callbacks
- ImplementEnumType
- ImpementBitFieldType
- Enumerations
- Object Inheritance
- Reads C++ `include` headers and code files
Generates
- Human friendly CSharp framework
- CSharp enumerations
- Converts C++ object types to native CSharp types
- Builds CSharp classes using inheritance mimicking the C++ structure.
- Typed member properties
- Typed function parameters
- Pre-defined Overrides and Virtual member function typing
- Maps all class functions to CSharp objects (overriding and new`ing as necessary)
- Object creation helpers (For say creating a new ScriptObject, Player, etc)
- pInvoke generation (Mapping CSharp definitions to C++ definitions.)
General Fixes
- Builds on Win32/64 so you can build it as a native 64 bit application.
- Win32 build runs on Mono, so it should port to other mono platforms easily.
- OpenGL support
- New font support to support all fonts
- Added/updated all the Shaders and added in the ability to do compiled Shaders.
- Shader system updates and performance improvements
- DirectX 9 enhancement
- Bullet (Full integration)
- New project generator so it will create all projects types from 2008 to 2013 (we are still debugging the 2013 visual studio build)
- Polyhedral collision detection
- GMK to it stock.
- Tool rewrite (Pop outs, fixes, etc.)
- IPS Pro integration
- Advanced UI Kit integration
- Full Linux version (2014 Q2)
- Full Android version (2014 Q4)
- Can run in Hybrid Mode. (TorqueScript Only, TorqueScript/CSharp, CSharp only)
General C++ Bug Fixes and improvements
- All callbacks use new Callback system
- Callback System has been normalized
- Replacement of many obsolete `ConsoleFunction` and `ConsoleMethod` to their respective DefineConsoleFunction syntax.
- U32/S32 realignment (Consistency and Sanity checks)
- Replacement of T3D Standard class types with C++ Standard classes when possible.
- full support form Windows 8 tablets with touch interface (Q1 2014)
- full support form Ubuntu tablets with touch interface (Q2 2014)
- GuideBOT integration
- DirectX 11 renderer (Q4 2014) * GPGPU support, tessellation support, multi-threading support, Shader Model 5.0, s unified vertex and pixel shaders and Geometry Shaders
- OpenGL 4.x renderer
Now, on to your questions!
I am wondering about the usage of DNT, how does someone go at moving a project over from TorqueScript to C#?
It isn`t an easy endeavor, especially if it is your first attempt. If you use a lot of namespaces, super-classes, etc. it can be quite tricky to translate it to CSharp. There are features which simplify the process but a solid understanding of CSharp (and inheritance and object casting) is a requirement. I have completely converted the Starter example at least 4 or 5 times. The last time I converted it, it took about a week. The first time I converted it, it probably took me 2 months. (On the plus side, you do get priority email and forum support with your purchase and we usually get back to you within 24 hours or less). And usually, if I can`t resolve it over the Forums or Email, I will remote to your development machine and look over your code. Plus I plan to have a SDK with examples on release.
The main thing that has really spiked my interest is the cross-platform support
Yes, our goal is to have a platform where you can build your game on MS Windows and then just recompile it on a supported platform and you will be able to release your game with little extra work. Currently, everything we write is MONO/Linux compatible and we do have a full Client/Server version working of DNT(OMNI) running on Linux. We are working very hard at pushing out an Android and MS Surface build.
and the boost in performance.
In regards to performance, OMNI performs reliably faster than stock T3D.
Under no load running an empty mission we see about a solid 21% increase in FPS and 16.21% increase in MSPF over stock.
Where OMNI really shines though is under load.
One of our load tests is bring up a zone with 300 AI in it. The AI are split into 2 or more teams and constantly attack each other. A dedicated server (Core 2.3ghz) runs this zone at about 40% CPU utilization. The dedicated clients run upward of 60+ fps.
When running the same load test as a stand-alone we can usually run about 100 AI in the zone at 30-80 FPS.
Our test machine is an old Quad Core 2.3 Ghz dell computer, average AMD video card with 1 gig of ram.
Running with all Shaders turned on is also day and night between OMNI and stock T3D in regards to frame rate and MSPF.
The things that hold me back is that it`s a branch away from the standard T3D MIT and the third-party add-ons will need to be converted over to C#. Or can it work with a mixture of TorqueScript and C#?
OMNI can run TorqueScript, but yes, in general the resources need to be converted to CSharp.
We are working on incorporating the open source resources like GMK, Path Finding, etc. at launch, 3rd Party add-ons will take conversion.
In regards to patching of MIT fixes, we have a full time programmer who merges in the changes from the committee into our build. So at the very most we are a week or so behind the MIT build. Our code base at places is significantly different than stock MIT, but we try to evaluate each MIT patch for it`s worth and application. Some things done in the MIT branch just don`t make since, or are implemented poorly/incomplete. We fix these things by ensuring that networking/tools/etc. are upgraded to when the MIT patch misses one of them.
It would be great if I was able to move over to DNT, C# is a great language to use and the performance as I said really makes it standout to me as the correct way to go with T3D.
We couldn`t agree more.
One of the things that always killed me with stock T3D was that TorqueScript is proprietary. I cannot use any code I write in TorqueScript for anything else other than T3D. I have written tons of CSharp classes and code snippets over the years and it`s nice to know that I can just cut and paste them into the code w/out worrying about converting the syntax.
One other advantage we have noticed is that programmers working on our games are more apt to learn CSharp than TorqueScript. The CSharp language is a tool that can be used outside of the game development world, So learning it isn`t considered wasted time.
One final note on this is also the plethora of CSharp code available on the Internet. You can pretty much find code to do anything written in CSharp. It also simplifies things like database integration and 3rd party drivers (For use in things like robotics). If there is a CSharp driver, you`re in business. It opens up a whole new world to T3D, one where the limitations of engine integration no longer matter.
Vince Gee
Winterleaf Entertainment L.L.C.
About the author
www.winterleafentertainment.com
#2
There isn't much more to say about the Linux and Android build. Dusan is working very hard at getting the performance to where we want it. We have quite high standards at Winterleaf, and if the build doesn't run as well as the Windows version we kick it back for more refinements.
Our goal though of course is to provide a unified development experience where a game developer can write the game on there platform of choice and then just recompile the game to the different OS's w/out having to re-write anything.
OMNI, will be our flagship product. Our goal is to keep it actively developed, enhanced and patched. It will be not only a solid game engine but also a great serious game engine (Simulation) as well.
We will be building addons for OMNI, like an XMPP chat server, AI plug ins, etc. The idea is that each of the plug-ins will be certified to work with the OMNI engine and have a bolt on friendly integration.
Of course all of this will pivot around the C# language, taking advantage of what it gives us.
Imagine if you will, a game development environment where you have TRUE intellesense. Where you can switch between the C++ and the script in the same project. Where you can set break points in the C# and it will step into the C++. A real integrated development environment!
Then toss in all the improvements, improvements to the rendering, shaders, tools, script language, etc. All of this at your finger tips. Then if all of this wasn't enough, imagine SUPPORT! Got a problem, can't find the answer on the forums, email us! Get a person who is experienced in that part of the engine to answer your questions.
Got a question about DirectX/OpenGL?
Got a C# question?
Got a T3D question?
And so on. We might not have the exact answer, but we can definitely put you in the right direction and help you solve it.
Really the #1 reason why you should pick OMNI? Well, it's built off of T3D, it's an engine you are familiar with, you need to extend the c++ for your needs, you don't need to learn a new engine.
Price Point is still not decided. You see, we actually are paying people to implement some of this stuff. We have operational costs, support costs etc.
But we need to generate enough revenue to be able to keep our developers paid, and we would like to pay more and more people.
Right now we have been supporting our development costs by pumping EVERY dollar we make in sales into paying the developers. Paul and I take not a single penny from the operation. In fact we contribute a great deal of our own money into this operation on a monthly basis.
So price point? I don't know.
Will we charge per OS licensed? I don't know.
I do know one thing though....
We have already decided to offer a free version of the compiled Windows version of the C++ engine. Where people can modify the C# code to make there own game. Kinda like the old TGE days where you could build a game w/ the exe using scripts. So there will be plenty of ways to try the engine before purchasing it. There will be more on this coming in the future.
10/15/2013 (6:48 am)
@Jolinar,There isn't much more to say about the Linux and Android build. Dusan is working very hard at getting the performance to where we want it. We have quite high standards at Winterleaf, and if the build doesn't run as well as the Windows version we kick it back for more refinements.
Our goal though of course is to provide a unified development experience where a game developer can write the game on there platform of choice and then just recompile the game to the different OS's w/out having to re-write anything.
OMNI, will be our flagship product. Our goal is to keep it actively developed, enhanced and patched. It will be not only a solid game engine but also a great serious game engine (Simulation) as well.
We will be building addons for OMNI, like an XMPP chat server, AI plug ins, etc. The idea is that each of the plug-ins will be certified to work with the OMNI engine and have a bolt on friendly integration.
Of course all of this will pivot around the C# language, taking advantage of what it gives us.
Imagine if you will, a game development environment where you have TRUE intellesense. Where you can switch between the C++ and the script in the same project. Where you can set break points in the C# and it will step into the C++. A real integrated development environment!
Then toss in all the improvements, improvements to the rendering, shaders, tools, script language, etc. All of this at your finger tips. Then if all of this wasn't enough, imagine SUPPORT! Got a problem, can't find the answer on the forums, email us! Get a person who is experienced in that part of the engine to answer your questions.
Got a question about DirectX/OpenGL?
Got a C# question?
Got a T3D question?
And so on. We might not have the exact answer, but we can definitely put you in the right direction and help you solve it.
Really the #1 reason why you should pick OMNI? Well, it's built off of T3D, it's an engine you are familiar with, you need to extend the c++ for your needs, you don't need to learn a new engine.
Price Point is still not decided. You see, we actually are paying people to implement some of this stuff. We have operational costs, support costs etc.
But we need to generate enough revenue to be able to keep our developers paid, and we would like to pay more and more people.
Right now we have been supporting our development costs by pumping EVERY dollar we make in sales into paying the developers. Paul and I take not a single penny from the operation. In fact we contribute a great deal of our own money into this operation on a monthly basis.
So price point? I don't know.
Will we charge per OS licensed? I don't know.
I do know one thing though....
We have already decided to offer a free version of the compiled Windows version of the C++ engine. Where people can modify the C# code to make there own game. Kinda like the old TGE days where you could build a game w/ the exe using scripts. So there will be plenty of ways to try the engine before purchasing it. There will be more on this coming in the future.
#4
DNT has an open-source version; would that be compatible with OMNI's scripting? Where I'm going with this is I'd be really excited to develop code in an open-source engine, and build it into a proprietary one for deployment.
Also, using IronPython with T3D sounds fantastic if it has as tight integration with the .NET framework as C# sounds like it does. C# does not float my boat, but being able to do (and teach) desktop gamedev with Python certainly would.
10/15/2013 (4:45 pm)
Holy crap guys. This is fairly fantastic. The thing GG have always said about T3D going MIT is that if you wanted to, you could rebrand the engine and sell it yourself... it seems like someone's doing it, and doing a damn fine job if you can deliver on all of these features! I'd be so keen to jump on the bandwagon, but I've fallen in love with developing in an open-source environment. I don't think I'd enjoy moving back into a proprietary engine!DNT has an open-source version; would that be compatible with OMNI's scripting? Where I'm going with this is I'd be really excited to develop code in an open-source engine, and build it into a proprietary one for deployment.
Also, using IronPython with T3D sounds fantastic if it has as tight integration with the .NET framework as C# sounds like it does. C# does not float my boat, but being able to do (and teach) desktop gamedev with Python certainly would.
#5
I was careful only posting the features already complete. Between Dusan and I we have managed to roll all of it in. Now we are fixing and debugging.
Yes, you can use any DotNet language to use w/ OMNI or DNT. It's just syntax. So yes, you could use Iron Python etc.
OMNI's scripting interface is completely different than DNT. I spent a couple months going over all the code, optimizing, rewriting, improving. The end result was an integrated console so the C++ flows to the CSharp and vice-versa much smoother now.
Paul and I are still discussing licensing. I've always been the person who wants to give it away free and Paul has always been Scrooge McDuck.
Like always, we will find a middle ground where Indies can flourish and yet we can make enough money to continue engine development.
10/15/2013 (4:52 pm)
@Daniel,I was careful only posting the features already complete. Between Dusan and I we have managed to roll all of it in. Now we are fixing and debugging.
Yes, you can use any DotNet language to use w/ OMNI or DNT. It's just syntax. So yes, you could use Iron Python etc.
OMNI's scripting interface is completely different than DNT. I spent a couple months going over all the code, optimizing, rewriting, improving. The end result was an integrated console so the C++ flows to the CSharp and vice-versa much smoother now.
Paul and I are still discussing licensing. I've always been the person who wants to give it away free and Paul has always been Scrooge McDuck.
Like always, we will find a middle ground where Indies can flourish and yet we can make enough money to continue engine development.
#6
Bite my Scrooge McDuck. LOL
@Daniel
So let me ask you a question then. Whats an open-source environment to you? How is it that you feel that OMNI would be missing it?
Vince is right. we are still discussing licensing and all. We are looking at a "free" version, Indie version, Pro version, etc.... comparing pricing and the whole nine yards. Its a lot to take into consideration atm but our GOAL is to keep this "indie friendly".
10/15/2013 (5:10 pm)
@VinceBite my Scrooge McDuck. LOL
@Daniel
So let me ask you a question then. Whats an open-source environment to you? How is it that you feel that OMNI would be missing it?
Vince is right. we are still discussing licensing and all. We are looking at a "free" version, Indie version, Pro version, etc.... comparing pricing and the whole nine yards. Its a lot to take into consideration atm but our GOAL is to keep this "indie friendly".
#7
THAT LIST IS JUST TOO MUCH BIG TO IGNORE.
i think that is a new engine rather then t3d.
u guys just using existing structure to replace/integrate different open source product into it.
i am interested on android and mono part.always wanted to try out myself to integrate them.but time and my knowledge base is a problem.
lucky u are to get DUSHAN,the vim.
Anyway,like before again asking,
is there any plan to make mono part publicly available(in any chance)?
that way u guys also will be able to sell your c# add-ons for public.
10/15/2013 (8:27 pm)
SHIT!THAT LIST IS JUST TOO MUCH BIG TO IGNORE.
i think that is a new engine rather then t3d.
u guys just using existing structure to replace/integrate different open source product into it.
i am interested on android and mono part.always wanted to try out myself to integrate them.but time and my knowledge base is a problem.
lucky u are to get DUSHAN,the vim.
Anyway,like before again asking,
is there any plan to make mono part publicly available(in any chance)?
that way u guys also will be able to sell your c# add-ons for public.
#8
WOW ... I seriously need to look at moving Re-Spawn across to this one day. One step at a time. :)
10/15/2013 (11:09 pm)
@Paul: I might have missed it, but what would the price be for existing customers of DNT?WOW ... I seriously need to look at moving Re-Spawn across to this one day. One step at a time. :)
#9
"u guys just using existing structure to replace/integrate different open source product into it."
not quite. while yes, we are adding a few open source items in, its the ones that we feel should have been added long ago. HOWEVER, that being said, we are also adding alot of our own code to fix what we consider to be issues with the engine.
the Mono build is part of OMNI. we cant rip out that code base and sell it seperately, there are just too many changes that were done. however, once OMNI hits the market, why yes, it will be publically available.
@Quinton
We havent decided upon a price yet for existing DNT customers, but yes, there will be an upgrade path for you guys.
10/16/2013 (3:44 am)
@ahsan"u guys just using existing structure to replace/integrate different open source product into it."
not quite. while yes, we are adding a few open source items in, its the ones that we feel should have been added long ago. HOWEVER, that being said, we are also adding alot of our own code to fix what we consider to be issues with the engine.
the Mono build is part of OMNI. we cant rip out that code base and sell it seperately, there are just too many changes that were done. however, once OMNI hits the market, why yes, it will be publically available.
@Quinton
We havent decided upon a price yet for existing DNT customers, but yes, there will be an upgrade path for you guys.
#10
You have got my interest!
I do have a few questions however about the Omni engine.
Feature Questions -
1. What rendering technique are you using for the OpenGL 4 Renderer for OMNI? (Deferred Rendering? Physically Based Rendering?)
2. Will there be proper texture blending?
3. Will you guys be supporting SteamOS? I know that the spec is not out for Steam OS yet. However I just thought that I would ask.
4. Will the terrain atlas making a come back? or Will there be another tool for multiple terrains?
When is Omni expected to launch and how can I keep up to date on news from winter leaf?
10/27/2013 (8:27 pm)
@Winter Leaf Entertainment -You have got my interest!
I do have a few questions however about the Omni engine.
Feature Questions -
1. What rendering technique are you using for the OpenGL 4 Renderer for OMNI? (Deferred Rendering? Physically Based Rendering?)
2. Will there be proper texture blending?
3. Will you guys be supporting SteamOS? I know that the spec is not out for Steam OS yet. However I just thought that I would ask.
4. Will the terrain atlas making a come back? or Will there be another tool for multiple terrains?
When is Omni expected to launch and how can I keep up to date on news from winter leaf?
#11
I will ping Vince, sorry for the wait.
11/12/2013 (12:46 am)
@Benjamin, just bumping this since we haven't answered you yet!I will ping Vince, sorry for the wait.
#12
I definitely understand that you guys are a professional dev team, and need to make a living off this stuff - I don't begrudge you that at all! I don't subscribe to some of the more extreme views on software openness. But having tasted freedom I'd be reluctant to go back ;P.
11/12/2013 (1:35 am)
Sorry, just realised I didn't get back to you either! By 'open source environment', I mean the ability to fork the code, host it somewhere convenient like GitHub (without paying for a private repo), see what other people are doign (via the network diagram, or just by following individuals), submit patches and discuss them (which, to be fair, can be done in a closed ecosystem too), and, as we've discovered from T3D's new license, the freedom to take the code in whatever direction I choose (as you guys are now doing).I definitely understand that you guys are a professional dev team, and need to make a living off this stuff - I don't begrudge you that at all! I don't subscribe to some of the more extreme views on software openness. But having tasted freedom I'd be reluctant to go back ;P.
#13
I asked our Coding Guru Dusan the questions, and this is what I got back. Hope it helps,
Vince
1. We will try to add much as possible. Currently what I can say and how ( without discussing =] ) is to present new way of handling "basic/advanced" post-fx lighting inside OMNI.
Just to elaborate little this.
On technical side it is better that we transfer all of basic/advanced post-fx on card profiles and props. So low-end GPU's (Intel GMA series ...) on engine start automatically start with "minimum settings" - in OMNI , medium settings is reserved for mid-end GPU's (Radeon series 5xxx ..., GeForce 5xx, 6xx ... so on) while for "highest settings" we are planning something completely new. All settings can be overdrive via "advanced graphics options". Think about "highest settings" as "expert lighting" (you already have basic/advanced lighting inside T3D). All graphics settings (post processing, particles, shading, shadows, AA, motion blur, Anisotropic Filtering ... etc. ) can be changed inside it
*2. I don't know answer on that question. (Never checked that or know that isn't "properly" added)
3. SteamOS is how I understand just another fork of Linux (with custom desktop environment). I can just say that we will watch Valve and their work on SteamOS and see how people and companies will adopt whole SteamOS. At this moment, I don't see any reason why not to support SteamOS with other Linux distros.
*4. I don't know answer on that question.
All with questions with "*" I don't know answer
11/12/2013 (4:38 am)
@Benjamin,I asked our Coding Guru Dusan the questions, and this is what I got back. Hope it helps,
Vince
1. We will try to add much as possible. Currently what I can say and how ( without discussing =] ) is to present new way of handling "basic/advanced" post-fx lighting inside OMNI.
Just to elaborate little this.
On technical side it is better that we transfer all of basic/advanced post-fx on card profiles and props. So low-end GPU's (Intel GMA series ...) on engine start automatically start with "minimum settings" - in OMNI , medium settings is reserved for mid-end GPU's (Radeon series 5xxx ..., GeForce 5xx, 6xx ... so on) while for "highest settings" we are planning something completely new. All settings can be overdrive via "advanced graphics options". Think about "highest settings" as "expert lighting" (you already have basic/advanced lighting inside T3D). All graphics settings (post processing, particles, shading, shadows, AA, motion blur, Anisotropic Filtering ... etc. ) can be changed inside it
*2. I don't know answer on that question. (Never checked that or know that isn't "properly" added)
3. SteamOS is how I understand just another fork of Linux (with custom desktop environment). I can just say that we will watch Valve and their work on SteamOS and see how people and companies will adopt whole SteamOS. At this moment, I don't see any reason why not to support SteamOS with other Linux distros.
*4. I don't know answer on that question.
All with questions with "*" I don't know answer
#14
I look forward to using your engine.
Is there any way I stay up to date with the Omni Engine?
- Ben
11/12/2013 (4:58 pm)
@Vince Gee - Hey, thank you for the reply :)I look forward to using your engine.
Is there any way I stay up to date with the Omni Engine?
- Ben
#15
In the mean time, if you register at www.winterleafentertainment.com you will receive out newsletters which usually go out once a month covering things we are working on and planning on releasing.
Vince
11/12/2013 (5:18 pm)
We are currently working on a new website for OMNI, something that can showcase what we have been pulling off.In the mean time, if you register at www.winterleafentertainment.com you will receive out newsletters which usually go out once a month covering things we are working on and planning on releasing.
Vince
#16
:)
Do you guys have a rough price set for Omni? Ballpark Figure if anything?
I am guessing that It wont be free with all the work that gone into it.
11/12/2013 (5:49 pm)
@Vince Gee - Registered and Subscribed.:)
Do you guys have a rough price set for Omni? Ballpark Figure if anything?
I am guessing that It wont be free with all the work that gone into it.
#17
Where you can modify the C# but not the C++ for free. So assuming you didn't have to do anything special in C++ you could write a game for free using it.
I can't give a price yet on the SDK. This is the thing, we want to get the engine in as many hands as possible, but we also need to recover our investments to get the engine done and fund future development.
for example, right now we are fighting with tessellation. Soon we might have to discuss benching it for v.2 if we can't solve the problems. Well that will reduce the price, etc. Hopefully as December approaches we will have a better gauge of what the pricing will look like.
Vince
11/12/2013 (5:56 pm)
I don't know what the full source sdk price will be, but I do know we will have a free binary version.Where you can modify the C# but not the C++ for free. So assuming you didn't have to do anything special in C++ you could write a game for free using it.
I can't give a price yet on the SDK. This is the thing, we want to get the engine in as many hands as possible, but we also need to recover our investments to get the engine done and fund future development.
for example, right now we are fighting with tessellation. Soon we might have to discuss benching it for v.2 if we can't solve the problems. Well that will reduce the price, etc. Hopefully as December approaches we will have a better gauge of what the pricing will look like.
Vince
#18
11/12/2013 (6:06 pm)
Providing a free binary will be fantastic if it's paired with something like Jeff Raab's component system. Otherwise it'll be as hamstrung as binary-only T3D was.
#19
You know Winterleaf almost funded that project, we had talked to Jeff about it. I was hesitant on adopting it after realizing that each component has it's own simobject id and it would flood the sim dictionaries. Then there was the whole ghosting routines that I wasn't sold on.
I never heard whether he addressed my concerns and thus the reason we didn't move forward with it.
I loved the concept, but I wasn't sold on the implementation.
If this has changed, I'd be more than happy to consider putting something like that in, but as his solution standed then, it was not conducive to a 400+ actors in the same zone.
11/12/2013 (6:11 pm)
@Daniel,You know Winterleaf almost funded that project, we had talked to Jeff about it. I was hesitant on adopting it after realizing that each component has it's own simobject id and it would flood the sim dictionaries. Then there was the whole ghosting routines that I wasn't sold on.
I never heard whether he addressed my concerns and thus the reason we didn't move forward with it.
I loved the concept, but I wasn't sold on the implementation.
If this has changed, I'd be more than happy to consider putting something like that in, but as his solution standed then, it was not conducive to a 400+ actors in the same zone.
#20
As for a free Binary - Would that include all the graphical and technical features? Excluding source code of course. :)
On the subject of documentation - How deep/Indepth will the docs be?
Will there be global Illumination?
Sorry for all the questions you got me really excited now!
11/12/2013 (6:39 pm)
@Vince Gee - Tessellation! You really got my attention now.As for a free Binary - Would that include all the graphical and technical features? Excluding source code of course. :)
On the subject of documentation - How deep/Indepth will the docs be?
Will there be global Illumination?
Sorry for all the questions you got me really excited now!
J0linar
From all those
2 features - are waking my interest
can you elaborate on it a bit more
and what the plans are for OMNI
as in what price range are we talking here
after all - there are plenty of multiplatform Engines
and what makes OMNI so special and why should one pick it up?