TGE 1.4.2 - Could it be a significant asset for GarageGames?
by Kenneth Eves · 10/24/2012 (10:24 pm) · 15 comments
Offered as food for thought. Opinions are my own.
Garage Games has made a bold move into the open-source arena with release of T3D 1.2 under the MIT license. While I'm no insider, it seems to me that this direction opens up an advantage where T3D 1.2 is poised to make very large waves in game development because of the license putting the Garage Games organization in a unique position of providing services to developers needing consultant or development assistance services.
Although T3D is certainly the flagship engine for Garage Games it is only one of many holdings of Garage Games. Others include Torque 2D, Torque Game Engine Advance (formerly TSE), and Torque Game Engine which has existed in many versions.
Application of the same strategy used for T3D to the other engines seems a logical extension of the current program. However, due to license issues where third party software is bundled into some GG engines, those engines are unsuitable candidates for open-source license.
To my knowledge, one exception to the license issue exists in Torque Game Engine 1.4.2.
How could a six year old engine that is also 3D be considered a valuable asset when it overlaps functionality with the flagship T3D? Simply stated the two are, in my opinion, sufficiently different that they fit needs of different areas in game development.
142 is CPU dependent with little GPU load
T3D is balanced between CPU and GPU and performs poorly in a weak GPU environment
142 is primarily OpenGL
T3D is functional using Direct3D with OpenGL support lagging
142 has no shader support requirements
T3D does not function without shader support
142 has an extremely small memory footprint
T3D is, to my knowledge, much larger in memory
142 is functions well with small in-game environments but suffers in large
T3D is equally adaptable large or small
So add these up: CPU heavy, GPU light, OpenGL, no shader requirments, small in memory, ideal for small in-game environments.
1. To me that sounds ideal for casual games where players are not likely to be invested in performance hardware normally used by more devoted "gamers".
2. It also sounds ideal for web deployment where small games are quickly downloadable, sometimes rendered inside a browser. (see A Mystical Land using Portalarium plugin for browser capability for example. http://www.insidesocialgames.com/2011/05/24/a-mystical-land-mmog/)
3. And finally it sounds ideal as a transition point to portable casual gaming on platforms such as iPhone, iPad, as well as Android phones and tablets, all of which use OpenGL and/or derivatives for 3D rendering.
Of course, all of this is based on casual observation and I could easily have flaws in my understanding. However, in the case that GG is looking for direction into expanding the open-source licensing, as I see it, TGE 1.4.2 does seem to have potential and possibly very little in terms of competition.
Garage Games has made a bold move into the open-source arena with release of T3D 1.2 under the MIT license. While I'm no insider, it seems to me that this direction opens up an advantage where T3D 1.2 is poised to make very large waves in game development because of the license putting the Garage Games organization in a unique position of providing services to developers needing consultant or development assistance services.
Although T3D is certainly the flagship engine for Garage Games it is only one of many holdings of Garage Games. Others include Torque 2D, Torque Game Engine Advance (formerly TSE), and Torque Game Engine which has existed in many versions.
Application of the same strategy used for T3D to the other engines seems a logical extension of the current program. However, due to license issues where third party software is bundled into some GG engines, those engines are unsuitable candidates for open-source license.
To my knowledge, one exception to the license issue exists in Torque Game Engine 1.4.2.
How could a six year old engine that is also 3D be considered a valuable asset when it overlaps functionality with the flagship T3D? Simply stated the two are, in my opinion, sufficiently different that they fit needs of different areas in game development.
142 is CPU dependent with little GPU load
T3D is balanced between CPU and GPU and performs poorly in a weak GPU environment
142 is primarily OpenGL
T3D is functional using Direct3D with OpenGL support lagging
142 has no shader support requirements
T3D does not function without shader support
142 has an extremely small memory footprint
T3D is, to my knowledge, much larger in memory
142 is functions well with small in-game environments but suffers in large
T3D is equally adaptable large or small
So add these up: CPU heavy, GPU light, OpenGL, no shader requirments, small in memory, ideal for small in-game environments.
1. To me that sounds ideal for casual games where players are not likely to be invested in performance hardware normally used by more devoted "gamers".
2. It also sounds ideal for web deployment where small games are quickly downloadable, sometimes rendered inside a browser. (see A Mystical Land using Portalarium plugin for browser capability for example. http://www.insidesocialgames.com/2011/05/24/a-mystical-land-mmog/)
3. And finally it sounds ideal as a transition point to portable casual gaming on platforms such as iPhone, iPad, as well as Android phones and tablets, all of which use OpenGL and/or derivatives for 3D rendering.
Of course, all of this is based on casual observation and I could easily have flaws in my understanding. However, in the case that GG is looking for direction into expanding the open-source licensing, as I see it, TGE 1.4.2 does seem to have potential and possibly very little in terms of competition.
About the author
<3 TGE/A Quickly becoming a fan of T3D 1.2
#2
I think it would have value, but there are some serious class hierarchy flaws that were fixed between TGE and T3D. At any rate it should be staged later. I also think it is really probable that someone will adapt T3D to run on really low end hardware and completely replace the renderer. It is designed to do that from what I understand.
I think Lukas has very good point about 3D acceleration. I think someone said even some of the crappy built in GPUs are getting good enough to run the engine. Based upon the devalue rate of PC hardware people that are interested in gaming will have better hardware to run your game. Higher end games tend to drive demand for better hardware.
10/25/2012 (12:36 am)
Quote:If you open sourced both it might confuse new users, and development would be split between the two engines rather than having a single uniform Torque movement.I think this is a very valid point. Back in early TGEA days it seems the community split as well.
I think it would have value, but there are some serious class hierarchy flaws that were fixed between TGE and T3D. At any rate it should be staged later. I also think it is really probable that someone will adapt T3D to run on really low end hardware and completely replace the renderer. It is designed to do that from what I understand.
I think Lukas has very good point about 3D acceleration. I think someone said even some of the crappy built in GPUs are getting good enough to run the engine. Based upon the devalue rate of PC hardware people that are interested in gaming will have better hardware to run your game. Higher end games tend to drive demand for better hardware.
#3
I was kicking this around myself. I did a test that removed advanced lighting and a few of the Pixel Shader 3.0 shaders. Performance was decent.
It just seems a waste to re-invent the wheel when it already exists. TGE on Android would have a lot of potential. As it is, that's a license violation. I'm hoping this will change eventually.
10/25/2012 (12:57 am)
RE: I also think it is really probable that someone will adapt T3D to run on really low end hardware and completely replace the renderer.I was kicking this around myself. I did a test that removed advanced lighting and a few of the Pixel Shader 3.0 shaders. Performance was decent.
It just seems a waste to re-invent the wheel when it already exists. TGE on Android would have a lot of potential. As it is, that's a license violation. I'm hoping this will change eventually.
#4
10/25/2012 (6:06 am)
Don't think they can change that license violation. Dont think they can change the License's on TGE or they would made it free as stepping stone for people to buy T3d. 
#5
as far i have read bitGap use old hardware for their MMO(based on t3d).so i think they have found good success on that.
10/25/2012 (8:54 am)
"someone will adapt T3D to run on really low end hardware "as far i have read bitGap use old hardware for their MMO(based on t3d).so i think they have found good success on that.
#6
10/25/2012 (8:55 am)
It's our IP, we can change the license on it. We would need to do a thorough code investigation to ensure that we could release it under an open source license. Part of the Torque 3D refactor was based around removing dependencies on code that we did not own the rights to. 
#7
As for getting T3D running like TGE.. make a tiny terrain, user 128x128 or if you are feeling brave, add some 256x256 textures into it, nerf your poly count of each model character model down to 500 or so and you should be getting pretty close not just the performance but the look and feel of TGE too :)
10/25/2012 (12:33 pm)
I suggested that when T3D was released that TGE should have been released under GPL, I still think it should be released under GPL (*not* MIT for a variety of reasons...)As for getting T3D running like TGE.. make a tiny terrain, user 128x128 or if you are feeling brave, add some 256x256 textures into it, nerf your poly count of each model character model down to 500 or so and you should be getting pretty close not just the performance but the look and feel of TGE too :)
#8
Also, about TGE being more appropriate for smaller environments... Tribes 2 says hi ;). I do appreciate where you're coming from, but I agree with Lukas - I reckon it'd be more valuable to keep development effort focused on T3D. I wonder if there'd even be a way to port TGE's render layer to T3D, so we could keep the advantages of the new engine and so on.
10/25/2012 (1:34 pm)
Quote:(*not* MIT for a variety of reasons...)Sorry, you probably went through this when the release happened, but do you care to elaborate?
Also, about TGE being more appropriate for smaller environments... Tribes 2 says hi ;). I do appreciate where you're coming from, but I agree with Lukas - I reckon it'd be more valuable to keep development effort focused on T3D. I wonder if there'd even be a way to port TGE's render layer to T3D, so we could keep the advantages of the new engine and so on.
#9
10/28/2012 (9:33 am)
T3D does run well on modern low end GPUs. My macbook pro has an intel 3000 card (which is not even the latest intel integrated GPU) in it and I get 15 fps with AL on the T3D demos and 60 fps with Basic Lighting. Using Basic Lighting should still look better than TGE rendering. Also when it comes to mobile platforms, you are going to have to write an Opengl ES layer for either engine and pretty much all mobile hardware support opengl ES 2.0 now which requires shaders. My guess is the mobile platforms will be better optimized for shader based rendering. (purely a guess though)
#10
I was wrong on this one. The current OpenGL and OpenGL ES standards have dropped backwards compatibility with fixed function pipelines.
That means that current hardware with current drivers won't run 1.4.2. (confirmed for OpenGL ES, not 100% sure with Win / Linux / Mac )
1.4.2 might be useful by back-porting TMK. Also I'm not sure if TMK fully replaces the fixed function design.
OpenGL support didn't hit in TGEA until 1.8.0 (or so) which isn't useful in that there are no expectations of a TGEA open source.
So, as far as an Android port (what I was looking into)...
TGE 1.4.2 is too old.
TMK might make 1.4.2 useful.
TGE 1.5.0 to TSE to TGEA 1.8.2 are all locked up.
... so, if TMK can't or won't fix 1.4.2, then T3D is the only likely path for a port.
Thanks again for the feedback on the topic. Much appreciated!
11/02/2012 (5:36 am)
My follow up...I was wrong on this one. The current OpenGL and OpenGL ES standards have dropped backwards compatibility with fixed function pipelines.
That means that current hardware with current drivers won't run 1.4.2. (confirmed for OpenGL ES, not 100% sure with Win / Linux / Mac )
1.4.2 might be useful by back-porting TMK. Also I'm not sure if TMK fully replaces the fixed function design.
OpenGL support didn't hit in TGEA until 1.8.0 (or so) which isn't useful in that there are no expectations of a TGEA open source.
So, as far as an Android port (what I was looking into)...
TGE 1.4.2 is too old.
TMK might make 1.4.2 useful.
TGE 1.5.0 to TSE to TGEA 1.8.2 are all locked up.
... so, if TMK can't or won't fix 1.4.2, then T3D is the only likely path for a port.
Thanks again for the feedback on the topic. Much appreciated!
#11
11/02/2012 (5:48 am)
Torque 3D in Basic Lighting mode running in OpenGL on the Mac is pretty close to TGEA 1.8.x running in OpenGL on the Mac in what is going on under the hood.
#12
Let me provide some context to this by saying I am first, foremost and quite decidedly, an artist/creative type. I can kinda sorta understand code if I'm looking at it with someone explaining what it is to me. I was able to take some code Kenneth provided to "patch" the terrain bug in TGEA causing the "walls" of the terrain to glitch.. That's about the extent of it though. Not for a lack of trying. I've tried to learn programming... I have C++ books, Python books, Java books and such here, which I bought in a serious attempt to learn. And I tried, several times. I'm just not wired for it.
That said...
I've tried T3D a few times now and, while I love a number of the features it has in isolation, I just can not stand working with the terrain editor. It drives me nuts. In every (and I mean every) other terrain editor I work with, using the "smooth" tool attempts to smooth out rough edges, or to create a smoother transition between different elevations. In T3D, it acts like a strange sort of "flatten" tool combined with "lower terrain". It just doesn't work the way it should, intuitively, or based on the sort of "standard" established by other engines/editors - including TGE/TGEA. Flatten does a similar thing, it lowers the terrain rather than flattening it.
That's just a specific issue. There are other quirks with T3D that make it a frustrating experience for me. In all, I just find working with TGEA to be a more comfortable and even pleasant experience. Also, TGEA can still yield some awesome looking results, even by today's standards (it's all in the art assets) while maintaining great performance across a wide variety of machines.
Now, the reason I mentioned the "I'm not a programmer" thing is because if I intend to work on something using TGEA, I need to find someone who already owns a license themself. I can't "share" my license of it, and it's impossible to obtain a new license. So, it seems that without learning how to program myself, I'm SOL and have no choice but to use T3D (which is not a good option to me),or use another engine altogether. There are suitable options out there... I just find TGEA to be the most "comfortable" to work with, at least as an artsy/designer type.
So... if there were anything North of a snowball's chance in hell that TGEA could be made available, in some way, to people who'd like to use it in lieu of T3D or other options, that would be a great thing.
Just my two cents.
03/29/2013 (5:46 am)
I'd love to see TGE/A released as MIT, or otherwise open source. Hell, start selling it again for a reduced cost if that's what it takes.Let me provide some context to this by saying I am first, foremost and quite decidedly, an artist/creative type. I can kinda sorta understand code if I'm looking at it with someone explaining what it is to me. I was able to take some code Kenneth provided to "patch" the terrain bug in TGEA causing the "walls" of the terrain to glitch.. That's about the extent of it though. Not for a lack of trying. I've tried to learn programming... I have C++ books, Python books, Java books and such here, which I bought in a serious attempt to learn. And I tried, several times. I'm just not wired for it.
That said...
I've tried T3D a few times now and, while I love a number of the features it has in isolation, I just can not stand working with the terrain editor. It drives me nuts. In every (and I mean every) other terrain editor I work with, using the "smooth" tool attempts to smooth out rough edges, or to create a smoother transition between different elevations. In T3D, it acts like a strange sort of "flatten" tool combined with "lower terrain". It just doesn't work the way it should, intuitively, or based on the sort of "standard" established by other engines/editors - including TGE/TGEA. Flatten does a similar thing, it lowers the terrain rather than flattening it.
That's just a specific issue. There are other quirks with T3D that make it a frustrating experience for me. In all, I just find working with TGEA to be a more comfortable and even pleasant experience. Also, TGEA can still yield some awesome looking results, even by today's standards (it's all in the art assets) while maintaining great performance across a wide variety of machines.
Now, the reason I mentioned the "I'm not a programmer" thing is because if I intend to work on something using TGEA, I need to find someone who already owns a license themself. I can't "share" my license of it, and it's impossible to obtain a new license. So, it seems that without learning how to program myself, I'm SOL and have no choice but to use T3D (which is not a good option to me),or use another engine altogether. There are suitable options out there... I just find TGEA to be the most "comfortable" to work with, at least as an artsy/designer type.
So... if there were anything North of a snowball's chance in hell that TGEA could be made available, in some way, to people who'd like to use it in lieu of T3D or other options, that would be a great thing.
Just my two cents.
#13
03/29/2013 (6:54 am)
Huge chunks of the engine would have to be removed to avoid contractual property issues. And the pieces removed would leave you with an extremely broken engine that would either have to be fixed by the community or be rewritten using T3D's codebase, which would be a lot of work for little gain, unfortunately.
#14
A guy can wish, though.
What a mess that deal ended up to be.
Guess I'll have to look into other engines, then :-/
03/29/2013 (10:13 am)
Yeah, I think that's been mentioned before. Ownership issues and it's complicated, etc. etc.A guy can wish, though.
What a mess that deal ended up to be.
Guess I'll have to look into other engines, then :-/
Torque Owner Lukas Joergensen
WinterLeaf Entertainment
Almost any computer have a GPU these days, so sharing the work between gpu and cpu shouldn't be a problem.
If you open sourced both it might confuse new users, and development would be split between the two engines rather than having a single uniform Torque movement.
T3D doesn't have to be heavy on the computer, use basic lightning turn of PostFX use efficient models etc etc and it can run on low end hardware aswell.