3 years from now...
by Chris "Had Chris First" · in General Discussion · 11/10/2002 (11:08 am) · 53 replies
Since this community is mostly centered around Torque, then what will happen when Torque does become obsolete? Will this community slowly die? Will GG make another Torque? Or will we just put an "insert engine here" sign on the website, and wait for another engine we can adopt.
Thoughts?
Thoughts?
About the author
#2
But I'm fairly certain you're going to see at least one more engine show up here within the next year or so, based on the discussion I had with someone while at IGC. (Nope, not gonna really say much about it, mainly because I don't know all the details, don't know what they are planning for sure, or if it's even going to happen. Don't ask me. So there. ;-) GG is a pretty bright group as a whole - since Jeff T. doesn't seem to be looking at this as a short term concept (supporting Indie development) I'm pretty sure they already have some ideas where to move things to next, either with a new engine(s), or with different forms of support and comercialization of the Indie Games movement.
11/10/2002 (11:25 am)
There's a TON of engines out there. Right now, not even all the people on GG are using the Torque engine to start with. The community has value without the engine, but, it does provide a good rallying point.But I'm fairly certain you're going to see at least one more engine show up here within the next year or so, based on the discussion I had with someone while at IGC. (Nope, not gonna really say much about it, mainly because I don't know all the details, don't know what they are planning for sure, or if it's even going to happen. Don't ask me. So there. ;-) GG is a pretty bright group as a whole - since Jeff T. doesn't seem to be looking at this as a short term concept (supporting Indie development) I'm pretty sure they already have some ideas where to move things to next, either with a new engine(s), or with different forms of support and comercialization of the Indie Games movement.
#3
A huge chunk of the code is dedicated to lower-level manipulation of data / game objects which hasn't changed much in years. The networking is far superior to almost anything else on the market. The scripting engine is extremely powerful.
I suspect you are talking about graphics which is a tiny proportion of the engine compared to all the other functions. The bit you finally see is not the whole story but rather the icing on the cake.
The perception that the Torque is standing still is more of a problem than it actually being so.
The torque provides and I suspect will always provide the power for indie/professional developers to create extraordinary games, it just takes the talent, creativity and sheer determination.
The Torque engine will be around for a long time. If everyone else leaves then in ten years you will still find me burning updates into the CVS.
- Melv.
11/10/2002 (11:26 am)
Torque resources are continually being submitted. Code snipits of updates and techniques for creating games are also being submitted. The CVS system allows fixes and updates to be available instantly to the community.A huge chunk of the code is dedicated to lower-level manipulation of data / game objects which hasn't changed much in years. The networking is far superior to almost anything else on the market. The scripting engine is extremely powerful.
I suspect you are talking about graphics which is a tiny proportion of the engine compared to all the other functions. The bit you finally see is not the whole story but rather the icing on the cake.
The perception that the Torque is standing still is more of a problem than it actually being so.
The torque provides and I suspect will always provide the power for indie/professional developers to create extraordinary games, it just takes the talent, creativity and sheer determination.
The Torque engine will be around for a long time. If everyone else leaves then in ten years you will still find me burning updates into the CVS.
- Melv.
#4
For example, rather than three seperate teams working concurrently on bringing shaders to their own version of the engine, we pool resources and work together. Once it is ready, we make it available to the community as a resource.
The only price involved is that others do the same. :-)
Eric Forhan
Art Director
www.trajectoryzone.com
11/10/2002 (11:32 am)
There is a group of programmers and artists getting together from different teams for a common goal: bringing new features to the engine.For example, rather than three seperate teams working concurrently on bringing shaders to their own version of the engine, we pool resources and work together. Once it is ready, we make it available to the community as a resource.
The only price involved is that others do the same. :-)
Eric Forhan
Art Director
www.trajectoryzone.com
#5
People tend to overlook some of the cool things Torque does do. The animation system is state of the art, and the network code is really good.
AND,
using the Torque, we were able to do our Macintosh port in a little under three hours, and most of that time what spent figuring out OSX's project builder.
11/10/2002 (11:34 am)
Gotta agree with Melv on this one. The graphics end might need some updating, but it is not stopping us from making a game that looks good and plays well. People tend to overlook some of the cool things Torque does do. The animation system is state of the art, and the network code is really good.
AND,
using the Torque, we were able to do our Macintosh port in a little under three hours, and most of that time what spent figuring out OSX's project builder.
#6
I agree 100% with the "gameplay, not graphics" point of view that has been espoused by the GG members. Gameplay should come before technology, eye candy, and even audiovisual design quality (ie, you can have a great game even using placeholder or 'engineer' art).
However, I think this is a very legitimate concern. The main reason is that upgrading Torque graphically is a very, very daunting task, and I don't mean programmatically using the exxisting engine to its full potential (a la Melv and others' great work), I mean improving fundamental things.
What does the future have in store? We can't afford to write off pixel and vertex shaders wholesale; like it or not, they ARE the future of graphics, and they aren't even particularly cutting-edge any more. What about dynamic lighting on the terrain? What about dynamic shadowmaps -- volume-based self-shadowing, dynamic vertex-based, dynamic self-shadowing vertex-based (hmmmmmmm . . . has this been done yet? hell, maybe it's impossible), but whatever method we use, it's undeniable that the future (and increasingly, the present) of lighting is real-time, not 'baked'.
What about buildings, interiors -- when will we move away from Quark and go the route of UT2k3, and be able to import meshes (some with simplified collision meshes) from modelling apps for all of our buildings? Dynamic lighting (or at least the option of dynamic lighting) on ALL world geometry?
There are so many things that need (or will need, within a year or two) to be updated and changed, that the solution may well be to recreate huge chunks of core Torque code completely from scratch.
To me, apart from all of the things that make Torque a great deal, it has 2 outstanding features that also make it a great engine: its beautiful netcode, and the indepth heightfield editing/mission editing abilities. However, as our means of creating content for games continues to change from ingame editors to external content creation programs, the usefulness of even this great, unique asset could be called into questions. Several games have come out recently with terrain that easily bests Torque's -- for instance, Battlefield 1942 -- and others are at the very least on par with it, for instance, ut2k3. Certainly, in each case, the elegance of the methods used to create the terrain could be called into question (particularly the bf1942 method, which works perfectly for the needs of that game but is VERY limited oterwise). But the end user will not see the difference. This will call for Torque to improve its terrain engine capability -- this is not necessary now, of course, but it will be eventually if we want Torque to remain a good-looking engine.
The netcode, however, is the best asset of Torque. There is no question whatsoever about that; Torque's netcode rules. Just from an end-user experience, it is definitely the best out there. UT2k3 is smooth enough for the number of players in the game, BF1942 is ABYSMAL, and Tribes 1 was great as well in the netcode department -- but Torque has them all trounced with its butter-like netplay.
The netcode and the terrain engine are also the #1 reason to upgrade Torque instead of starting over with better tech. Take those two things away, and what would you have? You'd have Half-Life with no multiplayer (but with vehicle physics and advanced animation capabilities). As it is, you've got a great engine saddled with a few sticking points -- sticking points that could really hinder the growth of the engine as a whole.
All of that being said, I still can't wait until someone really starts stretching Torque to its limits graphically. Melv has been doing an awe-inspiring job, and implementing his improvements fully can help make Torque look on par with many of the latest games that have come out. Certainly Torque has the capability to make a game that looks as good as, for example, Battlefield 1942, despite the added burden of creating so much game geometry with Quark. Currently, Torque has nothing to be ashamed of as an engine. But if a few important improvements aren't made, it will be very difficult to continue to keep people excited about upcoming games, or community projects like Realm Wars.
11/10/2002 (1:54 pm)
This really is an interesting point, and I think it does deserve some mention.I agree 100% with the "gameplay, not graphics" point of view that has been espoused by the GG members. Gameplay should come before technology, eye candy, and even audiovisual design quality (ie, you can have a great game even using placeholder or 'engineer' art).
However, I think this is a very legitimate concern. The main reason is that upgrading Torque graphically is a very, very daunting task, and I don't mean programmatically using the exxisting engine to its full potential (a la Melv and others' great work), I mean improving fundamental things.
What does the future have in store? We can't afford to write off pixel and vertex shaders wholesale; like it or not, they ARE the future of graphics, and they aren't even particularly cutting-edge any more. What about dynamic lighting on the terrain? What about dynamic shadowmaps -- volume-based self-shadowing, dynamic vertex-based, dynamic self-shadowing vertex-based (hmmmmmmm . . . has this been done yet? hell, maybe it's impossible), but whatever method we use, it's undeniable that the future (and increasingly, the present) of lighting is real-time, not 'baked'.
What about buildings, interiors -- when will we move away from Quark and go the route of UT2k3, and be able to import meshes (some with simplified collision meshes) from modelling apps for all of our buildings? Dynamic lighting (or at least the option of dynamic lighting) on ALL world geometry?
There are so many things that need (or will need, within a year or two) to be updated and changed, that the solution may well be to recreate huge chunks of core Torque code completely from scratch.
To me, apart from all of the things that make Torque a great deal, it has 2 outstanding features that also make it a great engine: its beautiful netcode, and the indepth heightfield editing/mission editing abilities. However, as our means of creating content for games continues to change from ingame editors to external content creation programs, the usefulness of even this great, unique asset could be called into questions. Several games have come out recently with terrain that easily bests Torque's -- for instance, Battlefield 1942 -- and others are at the very least on par with it, for instance, ut2k3. Certainly, in each case, the elegance of the methods used to create the terrain could be called into question (particularly the bf1942 method, which works perfectly for the needs of that game but is VERY limited oterwise). But the end user will not see the difference. This will call for Torque to improve its terrain engine capability -- this is not necessary now, of course, but it will be eventually if we want Torque to remain a good-looking engine.
The netcode, however, is the best asset of Torque. There is no question whatsoever about that; Torque's netcode rules. Just from an end-user experience, it is definitely the best out there. UT2k3 is smooth enough for the number of players in the game, BF1942 is ABYSMAL, and Tribes 1 was great as well in the netcode department -- but Torque has them all trounced with its butter-like netplay.
The netcode and the terrain engine are also the #1 reason to upgrade Torque instead of starting over with better tech. Take those two things away, and what would you have? You'd have Half-Life with no multiplayer (but with vehicle physics and advanced animation capabilities). As it is, you've got a great engine saddled with a few sticking points -- sticking points that could really hinder the growth of the engine as a whole.
All of that being said, I still can't wait until someone really starts stretching Torque to its limits graphically. Melv has been doing an awe-inspiring job, and implementing his improvements fully can help make Torque look on par with many of the latest games that have come out. Certainly Torque has the capability to make a game that looks as good as, for example, Battlefield 1942, despite the added burden of creating so much game geometry with Quark. Currently, Torque has nothing to be ashamed of as an engine. But if a few important improvements aren't made, it will be very difficult to continue to keep people excited about upcoming games, or community projects like Realm Wars.
#7
If Torque does not suit your needs, then feel free to make a game with another $100 competitor. If your game needs a $250,000 engine, then write the check.
Don't worry about things so much. From what I can see, the people most worried here are web developers or non-Torque owners that are not programmers. People that are actually making games, are just doing it.
Sorry for the rant, but I really can't take it any more.
Jeff Tunnell GG
11/10/2002 (2:28 pm)
I am getting pretty tired of these kinds of arguments. Just make your game. Don't worry about where we will be in three years. Don't worry about competing against games or engines that are not on the market.If Torque does not suit your needs, then feel free to make a game with another $100 competitor. If your game needs a $250,000 engine, then write the check.
Don't worry about things so much. From what I can see, the people most worried here are web developers or non-Torque owners that are not programmers. People that are actually making games, are just doing it.
Sorry for the rant, but I really can't take it any more.
Jeff Tunnell GG
#8
There's an effort to make new tech for Torque all the time - by people who are do'ers, not complainers. Shaders is being looked at, someone was planning on submitting DOT3 Bumpmapping recently, and I've been looking at things like building meshes in-game (I may need it for a future project.) Oddly enough, the original poster must either not be a Torque licensee, or, must not look around in the forums enough to see what people are seriously considering doing with the engine.
I don't think the Torque engine is the end-all be-all solution for game development when it comes to a 3D engine. But it's a damned good place to start from, and at it's price, you pretty much can't beat it. And throw in the Torque portion of the GG community here and you CAN'T beat it period. (Ok, there's some project that might be ill suited for it. MMORPG's come to mind.)
I'm kinda surprised at how often people say stuff like this anyway - again, doing some forums diving you can find a LOT of ways to improve the engine with little more than integrating some code and recompiling.
(I know - I'm just reitterating what a lot of people have said. My original post I was replyin' to the concept of "what's after this engine?".)
And you know... it just occured to me. We're Indie game developers. What the hell does three years from now matter anyway? We need to get games written, and released NOW, not three years from now. If the engine is truly outdated by then, and looks like a C64 in comparison with modern machines at the time, we'll all have to address the issue then. But until then, write some games, get them out there, and try like hell to make some money. Then if it IS an issue, and GG isn't able to provide... well, heck with it, license something else. But for now, it's all friggin' irrelevant. More game writin', less blue skyin'.
(Hehehe - normally I advise people to take a minute before replying to anything that could get ya' wound up. In this case, the more I thought about it, the more it wound me up ;-)
11/10/2002 (2:47 pm)
Ouch, Jeff - that was kind of harsh. Unluckly, there was also alot of truth to it. Those that make the most noise about the engine being dated don't doing anything about it. Those that think there is a problem and are working on games just fix the problem and move on (sometimes submitting it back as a resource, sometimes keeping the 'edge' for themselves.)There's an effort to make new tech for Torque all the time - by people who are do'ers, not complainers. Shaders is being looked at, someone was planning on submitting DOT3 Bumpmapping recently, and I've been looking at things like building meshes in-game (I may need it for a future project.) Oddly enough, the original poster must either not be a Torque licensee, or, must not look around in the forums enough to see what people are seriously considering doing with the engine.
I don't think the Torque engine is the end-all be-all solution for game development when it comes to a 3D engine. But it's a damned good place to start from, and at it's price, you pretty much can't beat it. And throw in the Torque portion of the GG community here and you CAN'T beat it period. (Ok, there's some project that might be ill suited for it. MMORPG's come to mind.)
I'm kinda surprised at how often people say stuff like this anyway - again, doing some forums diving you can find a LOT of ways to improve the engine with little more than integrating some code and recompiling.
(I know - I'm just reitterating what a lot of people have said. My original post I was replyin' to the concept of "what's after this engine?".)
And you know... it just occured to me. We're Indie game developers. What the hell does three years from now matter anyway? We need to get games written, and released NOW, not three years from now. If the engine is truly outdated by then, and looks like a C64 in comparison with modern machines at the time, we'll all have to address the issue then. But until then, write some games, get them out there, and try like hell to make some money. Then if it IS an issue, and GG isn't able to provide... well, heck with it, license something else. But for now, it's all friggin' irrelevant. More game writin', less blue skyin'.
(Hehehe - normally I advise people to take a minute before replying to anything that could get ya' wound up. In this case, the more I thought about it, the more it wound me up ;-)
#9
Torque may not be the most up to date in the graphics but I feel you can still do alot with it, and people are allthe time working on things like the bump mapping. Which I feel would help keep people looking at the game and playing it to be able to get the game play, which I feel we need to focus on.
11/10/2002 (5:25 pm)
I think that games are built on there game play, but many people just want graphics before really looking at the game play, or even sitting down to it.Torque may not be the most up to date in the graphics but I feel you can still do alot with it, and people are allthe time working on things like the bump mapping. Which I feel would help keep people looking at the game and playing it to be able to get the game play, which I feel we need to focus on.
#10
My concern isn't that Torque needs some kind of 'visual overhaul', or that it 'looks dated', or anything like that. All of my concerns are about the future of the engine; you have to admit, we haven't seen a lot of core graphical engine improvements yet. This might just be that it's going to take a while until we get dev teams that are large enough to have the resources, or the time, or the inclination, to make really impressive graphical upgrades. Or simply that there hasn't been enough time yet, and since the graphics are pretty spiffy thus far, no one feels that improving the engine graphically is worth the effort.
Of course, we have seen a lot of promise from the Legends team (bump mapping on terrain . . . maybe even DOT3 bump mapping?) and there are people adding pixel shader support to interiors as well, so maybe it's just a matter of time until these are fully workable, optimized and included into Torque. Again, future improvements, not any kind of perceived deficiency.
The one area, though, that could definitely be improved is the reliance on Quark/Hammer. That means adding a whole new skill necessity to a team; it isn't enough to have a coder god and a professional modeller -- you also need someone who can use an old level-editing tool. If Torque had the ability to import that kind of static geometry from the 3d modelling program, you could massively streamline the workflow -- not to mention that you would almost instantly boost the look of the Torque engine, simply because the content would be much higher quality.
That's probably not a small update, either. It would probably involve changing the way static geometry is lit (no more flat shading), not to mention the visibility calculation problems (since Quark handled that stuff for you). But I think everyone would agree that it's a very, very important feature to have, one that would make Torque easier to use, etc.
11/10/2002 (5:40 pm)
No, you misunderstand me! Graphically, the engine isn't dated -- now. But as I understand it (and I could be wrong), 'upgrading' the engine to support the latest and greatest graphical features is going to be pretty tough, for a number of reasons -- particularly any DirectX 8 or 9 features, since it's an OpenGL engine.My concern isn't that Torque needs some kind of 'visual overhaul', or that it 'looks dated', or anything like that. All of my concerns are about the future of the engine; you have to admit, we haven't seen a lot of core graphical engine improvements yet. This might just be that it's going to take a while until we get dev teams that are large enough to have the resources, or the time, or the inclination, to make really impressive graphical upgrades. Or simply that there hasn't been enough time yet, and since the graphics are pretty spiffy thus far, no one feels that improving the engine graphically is worth the effort.
Of course, we have seen a lot of promise from the Legends team (bump mapping on terrain . . . maybe even DOT3 bump mapping?) and there are people adding pixel shader support to interiors as well, so maybe it's just a matter of time until these are fully workable, optimized and included into Torque. Again, future improvements, not any kind of perceived deficiency.
The one area, though, that could definitely be improved is the reliance on Quark/Hammer. That means adding a whole new skill necessity to a team; it isn't enough to have a coder god and a professional modeller -- you also need someone who can use an old level-editing tool. If Torque had the ability to import that kind of static geometry from the 3d modelling program, you could massively streamline the workflow -- not to mention that you would almost instantly boost the look of the Torque engine, simply because the content would be much higher quality.
That's probably not a small update, either. It would probably involve changing the way static geometry is lit (no more flat shading), not to mention the visibility calculation problems (since Quark handled that stuff for you). But I think everyone would agree that it's a very, very important feature to have, one that would make Torque easier to use, etc.
#11
It's not because of any defects of Torque, either, it's just that I know I don't have time for this -- right now. Maybe in a month, or a year, or whatever, but I can't even *make* time for this. Again -- YET.
It's my goal to be able to work part-time and support myself. When that happens, I can indulge my game dev desires. I already have 2 major side activities that take up 80 and 32 hours per month, respectively, plus all of my hobbies -- this in addition to full-time work. So, Torqing off isn't an option for me now.
That's why I tend to only post in these kinds of threads. :) I may buy the engine next month, or next year, depending on how everything works out -- but in any event, it's in the future. So I am more interested in where Torque is going than where it is now. Although I do try to read as much pertinent stuff (tutorials etc) as I can slip in (like the scripting tutorials, testing out scripts in RW, messing around with the mapping stuff in RW occasionally, etc).
The major thing, tho, is having to use Hammer/Quark. :P I'm working on coding and 3d modelling -- learning another skill (and one that is REALLY inapplicable to any real-world jobs) is impossible for me to time-justify, and I'd hazard a guess that it's irritating for the guys out there that are actually making games right now. I mean, compare the effort that goes into a Quark building to the effort it would take to recreate same in Maya, Lightwave, or 3DSMax. Then think of how impossible it would be to recreate in Hammer/Quark some of the awesome, awesome things you can make in those programs.
11/10/2002 (5:57 pm)
Oh, ps -- heh, did I mention yet that I haven't licensed Torque? Well, I haven't. Yet.It's not because of any defects of Torque, either, it's just that I know I don't have time for this -- right now. Maybe in a month, or a year, or whatever, but I can't even *make* time for this. Again -- YET.
It's my goal to be able to work part-time and support myself. When that happens, I can indulge my game dev desires. I already have 2 major side activities that take up 80 and 32 hours per month, respectively, plus all of my hobbies -- this in addition to full-time work. So, Torqing off isn't an option for me now.
That's why I tend to only post in these kinds of threads. :) I may buy the engine next month, or next year, depending on how everything works out -- but in any event, it's in the future. So I am more interested in where Torque is going than where it is now. Although I do try to read as much pertinent stuff (tutorials etc) as I can slip in (like the scripting tutorials, testing out scripts in RW, messing around with the mapping stuff in RW occasionally, etc).
The major thing, tho, is having to use Hammer/Quark. :P I'm working on coding and 3d modelling -- learning another skill (and one that is REALLY inapplicable to any real-world jobs) is impossible for me to time-justify, and I'd hazard a guess that it's irritating for the guys out there that are actually making games right now. I mean, compare the effort that goes into a Quark building to the effort it would take to recreate same in Maya, Lightwave, or 3DSMax. Then think of how impossible it would be to recreate in Hammer/Quark some of the awesome, awesome things you can make in those programs.
#12
fyi, Jeffs right.
That bridge is far, so dont worry about crossing it until you are finished building your bridge (game)
metaphorically speaking of course.
11/10/2002 (6:25 pm)
Thanks for the Legends reference.fyi, Jeffs right.
That bridge is far, so dont worry about crossing it until you are finished building your bridge (game)
metaphorically speaking of course.
#13
11/10/2002 (6:34 pm)
Luc, I'm actually from the Legends dev team and did the bump mapping. Maybe you can refer me to a tutorial or a well documented example of dot3 bump mapping so I can actually put it in. I got my first taste of it with emboss, but dot3 sounds very interesting, and best of all, not that hard to convert from emboss. I've looked all through nVidia's and ATi's web site, gamasutra, google, and the like, but I can't find a tutorial or WELL DOCUMENTED example. I've found 3-4 demos w/ source, but nothing I can get any knowledge from.
#14
http://www.cfxweb.net/~delphigl/projects/dotbump.htm
11/10/2002 (7:02 pm)
All I've found thus far is an example with 'heavily documented source code'. The source is from a Delphi project, but the comments really are pretty extensive, and might help you understand what some of the other code samples. It *is* an OpenGL example . . .http://www.cfxweb.net/~delphigl/projects/dotbump.htm
#15
11/10/2002 (8:56 pm)
Well, it's documented, but it would help if it was in C++ :( 
#17
-J
11/11/2002 (12:23 am)
I wouldn't get too hung up on any post from the General Discussion Forum... anyone can post... and oh my, how they do...-J
#18
Does half life have these? No.
What is the most played FPS ever? Halflife.
The fact that people are making these things for the engine's graphics means that no, it isn't far behind on graphics, in fact, it's closer to the new stuff. While game graphics improve drastically every year, I think part of it is just more polygons + more hardware. The rate of innovation on graphics, engine-wise, is probably a bit slower than one would think. And still, people have shaders in Torque. They have bump mapping and per pixel lighting. In 3 years? Who knows what it will be.
--------
Oh sorry, I'm just pounding this into the heads of any doubters, I guess. Decided to post after reading only 40% of the thread.
11/11/2002 (12:29 pm)
Look at what people are talking about: dynamic lighting, mesh interiors, shaders, etc...Does half life have these? No.
What is the most played FPS ever? Halflife.
The fact that people are making these things for the engine's graphics means that no, it isn't far behind on graphics, in fact, it's closer to the new stuff. While game graphics improve drastically every year, I think part of it is just more polygons + more hardware. The rate of innovation on graphics, engine-wise, is probably a bit slower than one would think. And still, people have shaders in Torque. They have bump mapping and per pixel lighting. In 3 years? Who knows what it will be.
--------
Oh sorry, I'm just pounding this into the heads of any doubters, I guess. Decided to post after reading only 40% of the thread.
#19
11/11/2002 (1:26 pm)
Eric, that first one I already found. It's a good demo, and it's well documented, AND it's in C++. BUT, it uses quake 2 models for everything, and i have no idea how that stuff works :(
#20
Actually, John Carmack has said something like "I don't want more polygons, I want more texture passes".
There an article on 3dap a while back on the subject.
11/11/2002 (1:57 pm)
Quote:While game graphics improve drastically every year, I think part of it is just more polygons + more hardware.
Actually, John Carmack has said something like "I don't want more polygons, I want more texture passes".
There an article on 3dap a while back on the subject.
Eli McClanahan