Torque's 2D and 3D Future Vol. 2
by Brett Seyler · 11/18/2008 (6:01 am) · 76 comments
Since my last post on this subject in March, our Torque engine developers have been quietly toiling away on some major new stuff. The major focus? TGEA. While we've also spent some serious cycles on Torque X (the impressive results of which you'll be seeing soon) and on TGB (which you'll probably start hearing about toward the end of the year), probably 80% of the energy has gone to TGEA, and it's definitely starting to bear fruit.
There are many of you out there who started with TGE and have yet to make the jump to TGEA. Some of you will be asking "why TGEA and not TGE?" That's going to be the main focus of this post. I'm going to talk about what you'll be seeing from TGEA in the coming weeks and months, and why you're not likely to see an update to TGE soon, or perhaps every again. Don't panic yet. Keep reading...
While there were good reasons OpenGL support did not make it in TGEA 1.0, it's here now with vastly better performance than ever possible before thanks to the new Brian Richardson and Alex Scarborough powered graphic layer (GFX2). You can see the early fruits of Alex's efforts porting Rokkitball and Legions to the Mac for InstantAction below.
Rokkitball and Fallen Empire: Legions running in Safari on OSX
Did we mention that TGEA 1.8 is a FREE update for all TGEA owners? You can grab the beta release by hitting your My Account page right now.
What About TGE?
We've kicked this idea around a lot here in the office and so have you in the forums and in the comments on my previous blog. Why aren't we working on TGE? There are a number of reasons that matter to us, but I really want to approach this from your perspective. I want to take a hard look at the reasons people have, in the past, chosen TGE over TGEA.
Reason #1: TGE has better documentation. While this was true when TGEA 1.0 originally launched back in February 2007, it's certainly not anymore. TGEA documentation has gotten the Mich Perry treatment. It's now probably the most well-documented engine out there with mountains of resouces and a well organized set of reference and learning materials. See the proof for yourselves.
Mich's docs blogs:
* What do you think about Documentation?
* MAJOR Documentation Announcement 1
* New Documentation System
* Documentation Announcements
* September Documentation Wrap Up
* October Documentation Wrap Up
Reason #2: TGE has better resources. More resources? Yes. Better resources? Not anymore. TGEA is where all the latest, coolest stuff is happening, and some Torque developers are defintiely taking it by the horns and doing some really cool stuff. There are also a bunch of great add-ons for TGEA and many many more in the works. Case in point: Konrad Kiss. His Cliff Construction Kit makes gorgeous, detailed environments much easier to do. There's also Lorne McIntosh's SSAO (screen space ambient occlusion), and Apparatus' Beautiful Doors.
Reason #3: TGE has better stability. This is also no longer true. While TGEA pre-1.7 didn't match TGE's stability, the 1.7 and 1.7.1 release did. It may take a while for TGEA to match TGE's stability on the Mac, but it will get there. On the PC, it's no longer an issue.
Reason #4: TGE can run on OSX with OpenGL. Now so can TGEA, only it runs the same projects faster.
So with TGEA we now have...
The first thing I want to say is this: I'm thrilled we got here. With TGEA 1.8, we're out of reasons to look back and we'll be looking to the future from here with continued improvements to TGEA. It was not without great effort. TGEA is now is a much better product for game developers than TGE. If you're part of the Torque community, TGEA is where all the new development will be happening. TGEA is where the majority of new resources and 3rd party products will be born. And the best part of this for TGE owners is, we've taken great care to make sure that TGE projects are very portable to TGEA.
I'm encouraging those that haven't made the move to consider it. Having all hands on the same codebase benefits our development internally, but it also has huge benefits you guys. It means you'll have more compatible resources, and you'll have better, faster, more accurate answers to your technical questions because everyone's looking the same code.
There are real, tangible advantages for your project as well. In addition to all the hallmark strengths of Torque like world class networking and real-time world editing, with TGEA you have...
To us, it's a clear choice. You'll notice that we've taken this new embrace of Torque's future to heart with some recent changes to the website too. If you're a new user looking for an engine and want to make a 3D game, you'll end up on the TGEA landing page in just one click.
Getting to TGE is a tad tougher, but of course it's still there.
At some point, TGE will likely either go away, or be licensed in a different way. We could do something like the id does with the older Quake engines offering both GPL and closed source. We could simply leave it on the site as a less expensive option. To be honest, we haven't decided and we won't be doing anything about it right away.
This is a decision that I really would like some community feedback on. I know many of you have been working with TGE for years and really love it. We don't want to do anything that discourages it's continued use, but for new and advanced users alike, TGEA is now a better product, so we want to put it out front where we can focus on making it even better.
If TGE is going away soon, doesn't the name "TGEA" seem a little dumb?
Yes =) We are some gigantic dunces at this whole "naming" and "marketing" thing, but I see no reason to live with our mistakes or repeat them in the future. At some point, we'll look to simplify our products down to basics. It could look something like:
TGEA --> Torque 3D
and
TGB --> Torque 2D (duh...didn't we do that once already?)
So, none of that is happening *right* now, but likely happen in the future.
Okay Cool. So What's Next?
First off, to quickly review the stuff I've covered so far...
* TGEA 1.8 is in beta and adds an OpenGL graphics implementation and runs on OSX.
* At some point TGEA will replace TGE entirely and may be renamed to something more...concise.
Alright, alright! What's next? Good question...
The next version of Torque for 3D games
If you're still looking for another reason to make the leap to TGEA, pay attention. Remember all that stuff mentioned in my last blog under R&D? Well, back in March, there were dozens of things I wasn't ready to mention and quite a few that we'll be able to confirm as part of our next Torque release. The GG Studios team and many 3rd party developers we work with are already hands on with some of the coolest new R&D tech around. While the game studio is proving this new tech out with games you'll see on InstantAction and consoles over the next year, we're merging it with TGEA to be included in our next release.
Finally, Some "Thank Yous"
Maintaining and improving Torque is an incredible amount of work, so I want to recognize some of the people who've been HUGELY helpful in this last round of development, and generally moving the Torque experience in a positive direction, either by generously providing community resources, contributing directly to development, or providing really nice add-ons that help make Torque shine.
1. The entire Sickhead Games crew (Tom Spilman, Russell Fincher, Ross Pawley, and James Ford). These guys are to thank for a ton of what goes on behind the scenes in Torque development and even more of what you're about to see in the future. Their products, Torsion especially, make Torque much easier and much more fun to work with. They are a truly independent studio and they're bound for success.
2. John Kabus. John's continued support for the entire Torque product lineup has been a huge value add. He's been working on a lot of cool stuff that you guys will be psyched to see in the future too, so John definitely deserves a big thanks.
3. John Kanalakis. Remember when I mentioned that you'd be seeing some really new and impressive Torque X stuff? You have John to thank. John spearheaded our last round of development on Torque X both by updating it to XNA Game Studio 3.0 and he added his own suite of 3D editing tools. Now you can use WYSIWG tools to design your scene, assign components, and even write new ones right inside the editor. It's hot stuff and it's not been easy, so thank you, John. You're a superhero.
4. Melv May and Phillip O'Shea. Melv and Phillip are currently hard at work on some super-top-secret 2D engine stuff. When the next version of Torque's 2D engine tech ships, you'll have them to thank for most of its coolness.
5. Jeff Faust. Jeff never stops improving his fabulous AFX technology. He recently updated it to include TGEA support and more awesome effects. Word has it he's planning a big release in the next few months that will impress us all. I have no doubts about that. Keep it up, Jeff!
6. Rene Damm. Rene's another guy who's been absolutely killer behind the scenes helping with TGEA. He's tackled Atlas Terrain issues that frighten the best of us and his current plans include some really cool ways to leverage that core tech. We're all really looking forward to seeing what will happen when the final stuff emerges.
7. Laurentiu Danila a.k.a. "Apparatus". Does this guy ever sleep? Doors, bridges, props, textures, environments...KABOOM!. Talk about making Torque shine! Great job, Apparatus. Can't wait to see what you do next.
AND THAT'S A WRAP!
If you made it this far in the longest blog ever, congrats! I hope it was worth reading. Thanks!
...................................................................................................
Another Pre-emptive Q&A (with myself)
Q. What? No more TGE?
A. TGE is not gone yet and it's not going to disappear tomorrow, next week, or even next month. Next year? Possibly. There are no longer any technical reasons to choose TGE over TGEA, so we're going to focus our development efforts there. If you're concerned about the price, I wouldn't worry. We're going to find a way to make Torque technology as accessible on a budget as it can be. TGE resources and documentation aren't going to disappear either. They'll always be available, they may just become less relevant.
Q. Which of these releases will GG charge for?
A. TGEA 1.8 will be a free update. The next version of our 3D tech after that will be a paid update. The next version of our 2D tech will probably be a paid update. It's been almost 2 years since we charged for any version of TGEA and almost 4 years since we charge for any version of TGB. That's a pretty generous track record. At some point, we're going to have to feed the developers but as always, if we do charge for an update, you'll get more *way* more than your money's worth.
Q. When will we see OpenGL support in TGEA?
A. In TGEA 1.8! It's in open beta now! Pull it down. Put the hammer to it. Let us know what needs fixing.
There are many of you out there who started with TGE and have yet to make the jump to TGEA. Some of you will be asking "why TGEA and not TGE?" That's going to be the main focus of this post. I'm going to talk about what you'll be seeing from TGEA in the coming weeks and months, and why you're not likely to see an update to TGE soon, or perhaps every again. Don't panic yet. Keep reading...
While there were good reasons OpenGL support did not make it in TGEA 1.0, it's here now with vastly better performance than ever possible before thanks to the new Brian Richardson and Alex Scarborough powered graphic layer (GFX2). You can see the early fruits of Alex's efforts porting Rokkitball and Legions to the Mac for InstantAction below.
Rokkitball and Fallen Empire: Legions running in Safari on OSX
Did we mention that TGEA 1.8 is a FREE update for all TGEA owners? You can grab the beta release by hitting your My Account page right now.
What About TGE?
We've kicked this idea around a lot here in the office and so have you in the forums and in the comments on my previous blog. Why aren't we working on TGE? There are a number of reasons that matter to us, but I really want to approach this from your perspective. I want to take a hard look at the reasons people have, in the past, chosen TGE over TGEA.
Reason #1: TGE has better documentation. While this was true when TGEA 1.0 originally launched back in February 2007, it's certainly not anymore. TGEA documentation has gotten the Mich Perry treatment. It's now probably the most well-documented engine out there with mountains of resouces and a well organized set of reference and learning materials. See the proof for yourselves.
Mich's docs blogs:
* What do you think about Documentation?
* MAJOR Documentation Announcement 1
* New Documentation System
* Documentation Announcements
* September Documentation Wrap Up
* October Documentation Wrap Up
Reason #2: TGE has better resources. More resources? Yes. Better resources? Not anymore. TGEA is where all the latest, coolest stuff is happening, and some Torque developers are defintiely taking it by the horns and doing some really cool stuff. There are also a bunch of great add-ons for TGEA and many many more in the works. Case in point: Konrad Kiss. His Cliff Construction Kit makes gorgeous, detailed environments much easier to do. There's also Lorne McIntosh's SSAO (screen space ambient occlusion), and Apparatus' Beautiful Doors.
Reason #3: TGE has better stability. This is also no longer true. While TGEA pre-1.7 didn't match TGE's stability, the 1.7 and 1.7.1 release did. It may take a while for TGEA to match TGE's stability on the Mac, but it will get there. On the PC, it's no longer an issue.
Reason #4: TGE can run on OSX with OpenGL. Now so can TGEA, only it runs the same projects faster.
So with TGEA we now have...
The first thing I want to say is this: I'm thrilled we got here. With TGEA 1.8, we're out of reasons to look back and we'll be looking to the future from here with continued improvements to TGEA. It was not without great effort. TGEA is now is a much better product for game developers than TGE. If you're part of the Torque community, TGEA is where all the new development will be happening. TGEA is where the majority of new resources and 3rd party products will be born. And the best part of this for TGE owners is, we've taken great care to make sure that TGE projects are very portable to TGEA.
I'm encouraging those that haven't made the move to consider it. Having all hands on the same codebase benefits our development internally, but it also has huge benefits you guys. It means you'll have more compatible resources, and you'll have better, faster, more accurate answers to your technical questions because everyone's looking the same code.
There are real, tangible advantages for your project as well. In addition to all the hallmark strengths of Torque like world class networking and real-time world editing, with TGEA you have...
To us, it's a clear choice. You'll notice that we've taken this new embrace of Torque's future to heart with some recent changes to the website too. If you're a new user looking for an engine and want to make a 3D game, you'll end up on the TGEA landing page in just one click.
Getting to TGE is a tad tougher, but of course it's still there.
At some point, TGE will likely either go away, or be licensed in a different way. We could do something like the id does with the older Quake engines offering both GPL and closed source. We could simply leave it on the site as a less expensive option. To be honest, we haven't decided and we won't be doing anything about it right away.
This is a decision that I really would like some community feedback on. I know many of you have been working with TGE for years and really love it. We don't want to do anything that discourages it's continued use, but for new and advanced users alike, TGEA is now a better product, so we want to put it out front where we can focus on making it even better.
If TGE is going away soon, doesn't the name "TGEA" seem a little dumb?
Yes =) We are some gigantic dunces at this whole "naming" and "marketing" thing, but I see no reason to live with our mistakes or repeat them in the future. At some point, we'll look to simplify our products down to basics. It could look something like:
TGEA --> Torque 3D
and
TGB --> Torque 2D (duh...didn't we do that once already?)
So, none of that is happening *right* now, but likely happen in the future.
Okay Cool. So What's Next?
First off, to quickly review the stuff I've covered so far...
* TGEA 1.8 is in beta and adds an OpenGL graphics implementation and runs on OSX.
* At some point TGEA will replace TGE entirely and may be renamed to something more...concise.
Alright, alright! What's next? Good question...
The next version of Torque for 3D games
If you're still looking for another reason to make the leap to TGEA, pay attention. Remember all that stuff mentioned in my last blog under R&D? Well, back in March, there were dozens of things I wasn't ready to mention and quite a few that we'll be able to confirm as part of our next Torque release. The GG Studios team and many 3rd party developers we work with are already hands on with some of the coolest new R&D tech around. While the game studio is proving this new tech out with games you'll see on InstantAction and consoles over the next year, we're merging it with TGEA to be included in our next release.
Finally, Some "Thank Yous"
Maintaining and improving Torque is an incredible amount of work, so I want to recognize some of the people who've been HUGELY helpful in this last round of development, and generally moving the Torque experience in a positive direction, either by generously providing community resources, contributing directly to development, or providing really nice add-ons that help make Torque shine.
1. The entire Sickhead Games crew (Tom Spilman, Russell Fincher, Ross Pawley, and James Ford). These guys are to thank for a ton of what goes on behind the scenes in Torque development and even more of what you're about to see in the future. Their products, Torsion especially, make Torque much easier and much more fun to work with. They are a truly independent studio and they're bound for success.
2. John Kabus. John's continued support for the entire Torque product lineup has been a huge value add. He's been working on a lot of cool stuff that you guys will be psyched to see in the future too, so John definitely deserves a big thanks.
3. John Kanalakis. Remember when I mentioned that you'd be seeing some really new and impressive Torque X stuff? You have John to thank. John spearheaded our last round of development on Torque X both by updating it to XNA Game Studio 3.0 and he added his own suite of 3D editing tools. Now you can use WYSIWG tools to design your scene, assign components, and even write new ones right inside the editor. It's hot stuff and it's not been easy, so thank you, John. You're a superhero.
4. Melv May and Phillip O'Shea. Melv and Phillip are currently hard at work on some super-top-secret 2D engine stuff. When the next version of Torque's 2D engine tech ships, you'll have them to thank for most of its coolness.
5. Jeff Faust. Jeff never stops improving his fabulous AFX technology. He recently updated it to include TGEA support and more awesome effects. Word has it he's planning a big release in the next few months that will impress us all. I have no doubts about that. Keep it up, Jeff!
6. Rene Damm. Rene's another guy who's been absolutely killer behind the scenes helping with TGEA. He's tackled Atlas Terrain issues that frighten the best of us and his current plans include some really cool ways to leverage that core tech. We're all really looking forward to seeing what will happen when the final stuff emerges.
7. Laurentiu Danila a.k.a. "Apparatus". Does this guy ever sleep? Doors, bridges, props, textures, environments...KABOOM!. Talk about making Torque shine! Great job, Apparatus. Can't wait to see what you do next.
AND THAT'S A WRAP!
If you made it this far in the longest blog ever, congrats! I hope it was worth reading. Thanks!
...................................................................................................
Another Pre-emptive Q&A (with myself)
Q. What? No more TGE?
A. TGE is not gone yet and it's not going to disappear tomorrow, next week, or even next month. Next year? Possibly. There are no longer any technical reasons to choose TGE over TGEA, so we're going to focus our development efforts there. If you're concerned about the price, I wouldn't worry. We're going to find a way to make Torque technology as accessible on a budget as it can be. TGE resources and documentation aren't going to disappear either. They'll always be available, they may just become less relevant.
Q. Which of these releases will GG charge for?
A. TGEA 1.8 will be a free update. The next version of our 3D tech after that will be a paid update. The next version of our 2D tech will probably be a paid update. It's been almost 2 years since we charged for any version of TGEA and almost 4 years since we charge for any version of TGB. That's a pretty generous track record. At some point, we're going to have to feed the developers but as always, if we do charge for an update, you'll get more *way* more than your money's worth.
Q. When will we see OpenGL support in TGEA?
A. In TGEA 1.8! It's in open beta now! Pull it down. Put the hammer to it. Let us know what needs fixing.
About the author
Since 2007, I've done my best to steer Torque's development and brand toward the best opportunities in games middleware.
#22
11/18/2008 (2:30 pm)
I'm hyped about the future of TGEA. Even though it's sad that TGE is past its "best used by date" I feel it's the right direction for GG and the community to focus on TGEA (for 3D gaming) in order for us to keep competitive. 
#23
Marcus,
The answer to that is slightly complicated. Torque 3D (the next version of TGEA after 1.8) will include a lot of the R&D tech that we talked about back when we initially presented "Torque 2". In fact, TGEA 1.8 already includes some of that tech (GFX2, new ResourceManager, and String class). However, it won't include the "components" system that we talked about back then. That is still heavily in R&D. The base architecture for it is largely in place but we still need to develop a few games on it (keep an eye on InstantAction.com) and flesh out our component library. We also are hard at work on building the proper editor experience for it (and editors takes a long time to develop).
In a lot of ways TGEA 1.7 and TGEA 1.8 already have a lot of "modularization". With the Project Generator you can enable and disable different parts of the engine depending on what your goals are. We are slowly pushing this more and more with each iteration. I would expect it to start to "feel" a lot more modularized over the course of the next year or so.
In most practical ways, yes, the next version of TGEA (Torque 3D) is going to fill the role we envisioned for "Torque 2" (internally we've been calling it TGEA 2.0 for a while). There is still some cool tech in R&D waiting in the wings for you guys =)
11/18/2008 (3:04 pm)
Quote:
1) Is TGEA Torque 2?
2) Will TGEA be modularized like what was supposed to happen to Torque 2?
Marcus,
The answer to that is slightly complicated. Torque 3D (the next version of TGEA after 1.8) will include a lot of the R&D tech that we talked about back when we initially presented "Torque 2". In fact, TGEA 1.8 already includes some of that tech (GFX2, new ResourceManager, and String class). However, it won't include the "components" system that we talked about back then. That is still heavily in R&D. The base architecture for it is largely in place but we still need to develop a few games on it (keep an eye on InstantAction.com) and flesh out our component library. We also are hard at work on building the proper editor experience for it (and editors takes a long time to develop).
In a lot of ways TGEA 1.7 and TGEA 1.8 already have a lot of "modularization". With the Project Generator you can enable and disable different parts of the engine depending on what your goals are. We are slowly pushing this more and more with each iteration. I would expect it to start to "feel" a lot more modularized over the course of the next year or so.
In most practical ways, yes, the next version of TGEA (Torque 3D) is going to fill the role we envisioned for "Torque 2" (internally we've been calling it TGEA 2.0 for a while). There is still some cool tech in R&D waiting in the wings for you guys =)
#24
Funny... about two months ago I suggested that this was going to happen and got flamed...
http://www.garagegames.com/index.php?sec=mg&mod=resource&page=view&qid=15447
I got future sight! :)
Can you throw some numbers about the performance of TGEA on small or limited PCs using old processors (or Atom processors) and integrated cards? If it can beat TGE then I guess you have a winner.
Luck!
Guimo
11/18/2008 (3:36 pm)
Nice to know you have came to a decision and announced it!Funny... about two months ago I suggested that this was going to happen and got flamed...
http://www.garagegames.com/index.php?sec=mg&mod=resource&page=view&qid=15447
I got future sight! :)
Can you throw some numbers about the performance of TGEA on small or limited PCs using old processors (or Atom processors) and integrated cards? If it can beat TGE then I guess you have a winner.
Luck!
Guimo
#25
Someone correct me if I'm wrong (I get the versions and what features they do or don't have mixed up easily nowadays...).
11/18/2008 (3:46 pm)
@Andrew, well, that's one thing I haven't seen mentioned here and in the 1.8 beta postings. From what I recall, either 1.7 or an earlier (probably 1.7) release includes support for fixed function, which means likely TGEA would end up performing better in the cases of old hardware there as well.Someone correct me if I'm wrong (I get the versions and what features they do or don't have mixed up easily nowadays...).
#26
I ask only due the the current world financial problems, which inturn has major effects on exchange rates. Having almost saved money up to do the upgrade from TGE to TGEA, just before the big crash, I know how many others maybe feeling ATM. My personal example is that before the crash I was looking at @ $180 NZ dollars, but with the exchange rates it's now @ $300 NZ dollars, I'm sure others thinking ot doing the same thing are hurting ATM due to the same reason.
11/18/2008 (3:58 pm)
Matt I'd like to know how long the $145 Upgrade from TGE 1.5.2 to TGEA 1.7 will be available, and inturn the free upgrade to TGEA 1.8 from 1.7?I ask only due the the current world financial problems, which inturn has major effects on exchange rates. Having almost saved money up to do the upgrade from TGE to TGEA, just before the big crash, I know how many others maybe feeling ATM. My personal example is that before the crash I was looking at @ $180 NZ dollars, but with the exchange rates it's now @ $300 NZ dollars, I'm sure others thinking ot doing the same thing are hurting ATM due to the same reason.
#27
11/18/2008 (4:06 pm)
@Pierre, well I know you'd get 1.8 if you upgraded from TGE now. 1.8 is a free update as the GG guys here have said. 
#28
11/18/2008 (4:28 pm)
@Pierre - The TGE to TGEA upgrade will not be disappearing anytime soon and the upgrade from TGEA 1.7 to 1.8 will always be free. 
#29
I think a id approach would be nice. The problems I see with abandoning TGE, is TGEA is not a beginner's engine (I think thats even in the marketing material on the page) I guess beginners could always be pushed to Torque 2D. But if TGE existed in some way like the old Quake engines, beginners could still get their feet wet with it before coming to TGEA. And maybe have a note up front that the engines are "as is" with no updates, true support, etc.
Another, the upgrade price, or even the total price for TGEA. I know its still a hell of a deal at its current price, but some people just starting out might be a bit scared or some TGE users might not be able to afford the upgrade. But as Brett said, they are working on the price options.
All around though I'm excited about the future of TGEA and some of that cool R&D stuff.
11/18/2008 (4:34 pm)
I'm glad to see TGEA moving forward like it is, I'm enjoying it so much! Part of me still would hate to see TGE go, nostalgia and all :)I think a id approach would be nice. The problems I see with abandoning TGE, is TGEA is not a beginner's engine (I think thats even in the marketing material on the page) I guess beginners could always be pushed to Torque 2D. But if TGE existed in some way like the old Quake engines, beginners could still get their feet wet with it before coming to TGEA. And maybe have a note up front that the engines are "as is" with no updates, true support, etc.
Another, the upgrade price, or even the total price for TGEA. I know its still a hell of a deal at its current price, but some people just starting out might be a bit scared or some TGE users might not be able to afford the upgrade. But as Brett said, they are working on the price options.
All around though I'm excited about the future of TGEA and some of that cool R&D stuff.
#30
We honestly haven't had the time to do much profiling of TGE vs TGEA on older hardware. However, I do have some numbers from a test I did earlier this year that might be interesting. Before I get into that, though, here are a few things to think about:
How do we define "older" hardware?
Vista has been shipping on most OEM computers for close to 2 years now and it pretty much requires a shader model 2.0 video card. Even if you go back a little further than that, it was pretty common for OEM computers (like Dell, HP, Gateway, etc) to ship with an ATI or nVidia video card (possibly integrated). Anything the average user has bought in the last 5 or so years has been able at least as powerful as a GeForce 4 MX or a Radeon 1300. These aren't powerful cards by todays standards but they aren't slouches either. Anyone who is even remotely conscious of wanting to do 3D gaming is likely to have something at least as powerful as a Radeon 9600 (which was a *great* card btw).
If you look back at the GeForce 3 and Radeon 8500 (even some in the higher end GeForce 2's), there was a shift away from optimizing rendering performance by sending less data to the card (reducing the number of polys using software/cpu culling techniques) to being more intelligent about what order we send the data to the video card (worrying less about the number of polys and more about optimizing state and texture changes...in other words....batching).
TGE is heavily designed and built for pre-batching optimization. It's focus is on sending less triangles to the video card at the cost of using more cpu power. This was great for GeForce 2 and Voodoo 4's (the top-of-the-line video cards when the bulk of TGE's rendering code was written). However, with any "modern" video card (even weak ones), 300 single triangle drawcalls is way slower than one 3000 triangle drawcall (a batch). TGE uses a lot of OpenGL immediate mode draw calls and make very little use of deferred/buffered rendering (which can allow the video drivers to do a ton of optimization and to maximize use of you video bus). TGE just inherently draws slower on anything but the oldest or slowest of video cards.
TGEA will draw the same geometry faster on average on pretty much anything except those oldest and slowest cards and it may even win out there. But, you do have to be careful with TGEA...it opens you up to doing a lot more effects when you draw (specular, refraction, reflection, normal maps, etc) and it is very tempting to use them since they are so easy to use. If you overuse a lot of those effects it can really cost you in performance and TGEA does little to protect you from that.
How fast is a "small", "limited", or "old" processor on average these days? 1 GHz with a 512 Kb cache? Even the weakest of these is hugely faster than the processors that Torque was targeted against (Pentium 3 800 MHz). In addition to that, TGEA actually uses a lot less cpu power on average than TGE since we've offloaded a lot of the graphical effects (like the TerrainBlock blending and the dynamic shadows) from the cpu onto the gpu.
Now, before I paint us into a corner and have someone coming back with, "I tested it on my mom's machine and TGEA didn't run right but TGE did!", I do want to point out that most of the development on TGEA has been focused on newer and higher end video cards so there are bound to be some graphical glitches and issues when you run on older video cards. We've done a little patching here and there that helps (TGEA 1.7.1 plays a lot nicer with Intel integrated chips) but we know that there are still some gaps. It would be great to have it all work out of the box but for now we have higher priorities and, ultimately, it is a problem that will fix itself in time (what Clark Fagot calls the "elegant solution" =). It is our hope that the indies and the professional developers who actually ship games with Torque will be able to help us identify and fix these kinds of issues (we've already seen wins from some games on InstantAction.com). The more developers who join us on this codebase, the faster that will happen!
So back to those numbers I mentioned at the very beginning. Back in May I was doing some comparisons of the terrain rendering performance in the TGE Demo in TGE, TGEA 1.0.3, and TGEA 1.7. This is a comparison of the exact same scene in all three engines using the same artwork and effects. It was done on my 3 year old Dell laptop with an nVidia 8600 M (128 Mb video RAM):
Camera stationary:
TGE 200 fps
TGEA 1.0.3 180 fps
TGEA 1.7.0 320 fps
Camera in motion (repeating path):
TGE 180 - 250 fps
TGEA 1.0.3 180 - 240 fps
TGEA 1.7.0 300 - 360 fps
11/18/2008 (4:52 pm)
Guimo and Andrew,We honestly haven't had the time to do much profiling of TGE vs TGEA on older hardware. However, I do have some numbers from a test I did earlier this year that might be interesting. Before I get into that, though, here are a few things to think about:
How do we define "older" hardware?
Vista has been shipping on most OEM computers for close to 2 years now and it pretty much requires a shader model 2.0 video card. Even if you go back a little further than that, it was pretty common for OEM computers (like Dell, HP, Gateway, etc) to ship with an ATI or nVidia video card (possibly integrated). Anything the average user has bought in the last 5 or so years has been able at least as powerful as a GeForce 4 MX or a Radeon 1300. These aren't powerful cards by todays standards but they aren't slouches either. Anyone who is even remotely conscious of wanting to do 3D gaming is likely to have something at least as powerful as a Radeon 9600 (which was a *great* card btw).
If you look back at the GeForce 3 and Radeon 8500 (even some in the higher end GeForce 2's), there was a shift away from optimizing rendering performance by sending less data to the card (reducing the number of polys using software/cpu culling techniques) to being more intelligent about what order we send the data to the video card (worrying less about the number of polys and more about optimizing state and texture changes...in other words....batching).
TGE is heavily designed and built for pre-batching optimization. It's focus is on sending less triangles to the video card at the cost of using more cpu power. This was great for GeForce 2 and Voodoo 4's (the top-of-the-line video cards when the bulk of TGE's rendering code was written). However, with any "modern" video card (even weak ones), 300 single triangle drawcalls is way slower than one 3000 triangle drawcall (a batch). TGE uses a lot of OpenGL immediate mode draw calls and make very little use of deferred/buffered rendering (which can allow the video drivers to do a ton of optimization and to maximize use of you video bus). TGE just inherently draws slower on anything but the oldest or slowest of video cards.
TGEA will draw the same geometry faster on average on pretty much anything except those oldest and slowest cards and it may even win out there. But, you do have to be careful with TGEA...it opens you up to doing a lot more effects when you draw (specular, refraction, reflection, normal maps, etc) and it is very tempting to use them since they are so easy to use. If you overuse a lot of those effects it can really cost you in performance and TGEA does little to protect you from that.
How fast is a "small", "limited", or "old" processor on average these days? 1 GHz with a 512 Kb cache? Even the weakest of these is hugely faster than the processors that Torque was targeted against (Pentium 3 800 MHz). In addition to that, TGEA actually uses a lot less cpu power on average than TGE since we've offloaded a lot of the graphical effects (like the TerrainBlock blending and the dynamic shadows) from the cpu onto the gpu.
Now, before I paint us into a corner and have someone coming back with, "I tested it on my mom's machine and TGEA didn't run right but TGE did!", I do want to point out that most of the development on TGEA has been focused on newer and higher end video cards so there are bound to be some graphical glitches and issues when you run on older video cards. We've done a little patching here and there that helps (TGEA 1.7.1 plays a lot nicer with Intel integrated chips) but we know that there are still some gaps. It would be great to have it all work out of the box but for now we have higher priorities and, ultimately, it is a problem that will fix itself in time (what Clark Fagot calls the "elegant solution" =). It is our hope that the indies and the professional developers who actually ship games with Torque will be able to help us identify and fix these kinds of issues (we've already seen wins from some games on InstantAction.com). The more developers who join us on this codebase, the faster that will happen!
So back to those numbers I mentioned at the very beginning. Back in May I was doing some comparisons of the terrain rendering performance in the TGE Demo in TGE, TGEA 1.0.3, and TGEA 1.7. This is a comparison of the exact same scene in all three engines using the same artwork and effects. It was done on my 3 year old Dell laptop with an nVidia 8600 M (128 Mb video RAM):
Camera stationary:
TGE 200 fps
TGEA 1.0.3 180 fps
TGEA 1.7.0 320 fps
Camera in motion (repeating path):
TGE 180 - 250 fps
TGEA 1.0.3 180 - 240 fps
TGEA 1.7.0 300 - 360 fps
#31
I would say at this point that TGEA is more of a beginner engine than TGE. It has all of the same features and code that TGE plus better docs (which are also available offline as part of the SDK installer) plus an easier to understand directory structure plus better modularization plus bigger terrains plus an easier to use art pipeline (polysoup collision).
Honest question: where would you say that TGE is a better choice for a beginner at this point (aside from price)?
11/18/2008 (5:01 pm)
Chip,I would say at this point that TGEA is more of a beginner engine than TGE. It has all of the same features and code that TGE plus better docs (which are also available offline as part of the SDK installer) plus an easier to understand directory structure plus better modularization plus bigger terrains plus an easier to use art pipeline (polysoup collision).
Honest question: where would you say that TGE is a better choice for a beginner at this point (aside from price)?
#32
Honestly, I don't know why. I've seen a few people that just bought TGEA and not TGE complain about having some trouble with everything. Also I was basing it on this quote from the TGEA page:
There wasn't much of a learning curve for me, because I do believe the docs and everything is much better, but just by what I've seen, beginners seem to stumble through it, get frustrated and sometimes even ask for refunds.
Edit: I meant no offense or disrespect, and hope I didn't come like I did. :) I do think TGEA is much better and easier, hence why I upgraded and I think everyone should ;)
11/18/2008 (5:43 pm)
@MattHonestly, I don't know why. I've seen a few people that just bought TGEA and not TGE complain about having some trouble with everything. Also I was basing it on this quote from the TGEA page:
Quote:
TGEA is not a beginner's engine. You need to have a solid foundation for scripting, editing, physics, networking, animation, and debugging if you want to finish the game idea that's been stirring at the back of your mind for the last two years.
There wasn't much of a learning curve for me, because I do believe the docs and everything is much better, but just by what I've seen, beginners seem to stumble through it, get frustrated and sometimes even ask for refunds.
Edit: I meant no offense or disrespect, and hope I didn't come like I did. :) I do think TGEA is much better and easier, hence why I upgraded and I think everyone should ;)
#33
Thank you very much for the number which prove the architecture has indeed improved. Its undeniable that for a new project or maybe a beginner the superior codebase and documentation is the best choice they can make. It may have a higher price but its completely worth it. Even better if they go for the TGE-TGEA udpate path as it provides 2 engines for the price of one.
My question is there not because I want to run my game in my Gandmas PC. Unless you come from under a rock, netbooks are just selling like hot pancakes... their low price combined with decent hardware makes it an incredible choice for the average user. Come on... just look around you at the train or bus and you will see somebody with one of this tiny 10inch-800x600 books. And in this times of recession it is possible users in need of a portable PC will see them as a viable option.
Even major manufacturers and those who initially denied they wanted to enter this business have been driven into it forced by just the number of sales. Intel is still struggling to keep the demand for Atom and AMD in its worst move has delayed the launch of their new platform for 1 year at least.
So, you cannot deny that this new small form PCs are not the ones you find in your Grandpas library. These are brand new computers but with limited HW or dedicated video cards. Unfortunately you wont get an ATI 9600 (i really loved that card), or a nVidia FX5200 or a Geforce3Ti or any dedicated card in the foreseable future. We are talking about throwing up power for portability and the owners of these PCs will want games.
So, doesnt it makes sense for GG to keep development on TGE?
Dont take me wrong, Im completely in favor of a unified engine. All my game has been designed to be able to run in a unified network and it wont matter if the clients run in TGE-TGEA-TX-TWii-iTorque or whatever. Once finished I will start porting to TGEA in order to get an advanced client. But I just cannot justify a blind move to TGEA and just drop development on TGE if it will make me lose all this potential market. Indies live on their ability to reach any possible target and TGE gives us that.
So, please reconsider your idea of dropping TGE as it still has a lot of way to go.
Luck!
Guimo
11/18/2008 (5:49 pm)
Hi Matt,Thank you very much for the number which prove the architecture has indeed improved. Its undeniable that for a new project or maybe a beginner the superior codebase and documentation is the best choice they can make. It may have a higher price but its completely worth it. Even better if they go for the TGE-TGEA udpate path as it provides 2 engines for the price of one.
My question is there not because I want to run my game in my Gandmas PC. Unless you come from under a rock, netbooks are just selling like hot pancakes... their low price combined with decent hardware makes it an incredible choice for the average user. Come on... just look around you at the train or bus and you will see somebody with one of this tiny 10inch-800x600 books. And in this times of recession it is possible users in need of a portable PC will see them as a viable option.
Even major manufacturers and those who initially denied they wanted to enter this business have been driven into it forced by just the number of sales. Intel is still struggling to keep the demand for Atom and AMD in its worst move has delayed the launch of their new platform for 1 year at least.
So, you cannot deny that this new small form PCs are not the ones you find in your Grandpas library. These are brand new computers but with limited HW or dedicated video cards. Unfortunately you wont get an ATI 9600 (i really loved that card), or a nVidia FX5200 or a Geforce3Ti or any dedicated card in the foreseable future. We are talking about throwing up power for portability and the owners of these PCs will want games.
So, doesnt it makes sense for GG to keep development on TGE?
Dont take me wrong, Im completely in favor of a unified engine. All my game has been designed to be able to run in a unified network and it wont matter if the clients run in TGE-TGEA-TX-TWii-iTorque or whatever. Once finished I will start porting to TGEA in order to get an advanced client. But I just cannot justify a blind move to TGEA and just drop development on TGE if it will make me lose all this potential market. Indies live on their ability to reach any possible target and TGE gives us that.
So, please reconsider your idea of dropping TGE as it still has a lot of way to go.
Luck!
Guimo
#34
If I wait for the next big TGEA (Torque 3d) paid update, will it be still possible for me to upgrade from TGE?
11/18/2008 (6:01 pm)
The upgrade price is a bit expensive for some people (including me). As TGE is getting discarded (a great move btw), you'll need another less expensive option.If I wait for the next big TGEA (Torque 3d) paid update, will it be still possible for me to upgrade from TGE?
#35
11/18/2008 (6:27 pm)
But what about all the content packs? Many of the older packs will likely not get an update. Will developers update for TGEA? They will likely not sell much if they don,t.
#36
11/18/2008 (7:12 pm)
Its not like GG is going to stop selling TGE so the content packs are still going to sell. Also, the content packs should work fine in TGEA.
#37
You'll also find a large number of Vista shipped PC's are being downgraded to XP Pro. And also a Intel research project found no bonus from upgrading PC's to or with Vista on them. If anything there was more issues with security lock-downs and programmes crashing.
11/18/2008 (9:00 pm)
Matt you're right about Vista OEM shipping with "Named" branded PCs, like HP, Dell and others. But you'll find in the fine print that all have access install Windows XP. A few months back Microsoft stopped shipping media and licences to the big PC manufactors, but not to the small custom builder companys.You'll also find a large number of Vista shipped PC's are being downgraded to XP Pro. And also a Intel research project found no bonus from upgrading PC's to or with Vista on them. If anything there was more issues with security lock-downs and programmes crashing.
#38
11/18/2008 (11:42 pm)
Thanks for the updates. Can't wait to take everything out for a test run :)
#39
Brett I'm not wanting to side track your topic, but thought it would a good point to add.
So those who are able to help, please read and if possible put your hand up to offer any services you can offer.
The topic is covering updating some of the other GG development tools and content packs for use with the latest versions of TGE/A.
www.garagegames.com/mg/forums/result.thread.php?qt=80934
11/19/2008 (12:05 am)
Just a added note there is another topic running along side this topic. Brett I'm not wanting to side track your topic, but thought it would a good point to add.
So those who are able to help, please read and if possible put your hand up to offer any services you can offer.
The topic is covering updating some of the other GG development tools and content packs for use with the latest versions of TGE/A.
www.garagegames.com/mg/forums/result.thread.php?qt=80934
#40
Personal tastes and preferences aside, every game developer should be cheering Vista on!
Microsoft is one of the few movers in the computer industry that can establish shader capable cards as the baseline for all of our end users and that is incredibly exciting. Windows XP is a huge step backwards for game developers.
11/19/2008 (12:09 am)
Pierre,Personal tastes and preferences aside, every game developer should be cheering Vista on!
Microsoft is one of the few movers in the computer industry that can establish shader capable cards as the baseline for all of our end users and that is incredibly exciting. Windows XP is a huge step backwards for game developers.
Torque Owner Andrew Hull
How will a full switch to TGEA affect those trying to target older platforms?