Torque 2
by Martin Schultz · in Torque Game Engine · 10/11/2007 (7:29 am) · 75 replies
Hey guys,
interesting article over at Gamasutra about yesterday's introduction of InstantAction and esp. about what the Torque dev team says about the next evolution of Torque called "Torque 2".
From what Clark Fargot said I won't expect Torque 2 before mid 2008.
Anyway, interesting stuff.
Martin :-)
interesting article over at Gamasutra about yesterday's introduction of InstantAction and esp. about what the Torque dev team says about the next evolution of Torque called "Torque 2".
From what Clark Fargot said I won't expect Torque 2 before mid 2008.
Anyway, interesting stuff.
Martin :-)
Thread is locked
#3
Wont be for at least 2 years as I understood from the stream I watched. And I sure hope that the coming docs will be better than the TDN, but I did a lil jump when they mentioned that 'we might charge, for top level docs'.
No need to paint any scary scenarios, but I'm pretty sure this jump none the less will up the 'cost' for the choice, and that upped thingy might cut off a lot of the current customers/users of GG's products.
After all, the current range each targets very different ideas, and .... teamsize*
* being cash, manpower, knowledge, etc.
Anyways, very interesting stuff being presented!
10/11/2007 (8:02 am)
It is sure exiting to hear that they are considering a streamlined collective procuction ending up with a single engine. Might end up with a very cool package -engine-toolset-documentation that will be cross platform Win/Mac/Linux, OpenGL et all ;)Wont be for at least 2 years as I understood from the stream I watched. And I sure hope that the coming docs will be better than the TDN, but I did a lil jump when they mentioned that 'we might charge, for top level docs'.
No need to paint any scary scenarios, but I'm pretty sure this jump none the less will up the 'cost' for the choice, and that upped thingy might cut off a lot of the current customers/users of GG's products.
After all, the current range each targets very different ideas, and .... teamsize*
* being cash, manpower, knowledge, etc.
Anyways, very interesting stuff being presented!
#4
10/11/2007 (8:02 am)
Uhm, yeah, heard that name Juggernaut too as base of the new Torque 2 system that Clark talks about. Great thing btw. One common platform for all Torque flavours.
#5
It all sounded truly fantastic, but WAY out of scope for GG if history has taught us anything as we have waited and continue to wait on a truly release version of TGEA. The basics I took away from the future discussions is that TGE is dead, TGEA will be the "tge", and Torque2 will be the "new" TGEA (relatively speaking). No real timelines were discussed, and I also got the impression that all they have at this point is an idea and no real design at this point. We will most likely have to pony up for pre-release of T2 like TGEA owners did, and hope it gets some actual development focus before the technology its based on is depricated before its finished and Torque 3 discussions begin in earnest. I'm a bit jaded, but cooly optimistic that some real changes have taken place at GG. We will see.
10/11/2007 (8:10 am)
There was a live video stream of that part of the discussion yesterday afternoon (Thanks again, Dave). It all sounded truly fantastic, but WAY out of scope for GG if history has taught us anything as we have waited and continue to wait on a truly release version of TGEA. The basics I took away from the future discussions is that TGE is dead, TGEA will be the "tge", and Torque2 will be the "new" TGEA (relatively speaking). No real timelines were discussed, and I also got the impression that all they have at this point is an idea and no real design at this point. We will most likely have to pony up for pre-release of T2 like TGEA owners did, and hope it gets some actual development focus before the technology its based on is depricated before its finished and Torque 3 discussions begin in earnest. I'm a bit jaded, but cooly optimistic that some real changes have taken place at GG. We will see.
#6
10/11/2007 (8:15 am)
Quote:he IAC deal actually gives us the opportunity to sit back and rethink the foundation on which the tools are built, rather than hot-fixing bugs."Why dont I like this sentence?
#7
T2 is not going to happen anytime soon but I sure do hope that updates for TGE will.
10/11/2007 (9:32 am)
Quote:Why dont I like this sentence?Well, if you aren't a programmer then you won't. If you are then, well, you must be already pretty happy with the current state of the engine :)
T2 is not going to happen anytime soon but I sure do hope that updates for TGE will.
#8
No offense intended, but you are :) Torque 2 is already under active development (and has been for a few months now), and most importantly is using an iterative development process based off of our internal Juggernaut code base, moving forward to the new architecture (modules + components).
We are going to be using different web infrastructure from the official GG.com forums/web page for the transparent development program, so I'm not going to be kicking out tons of data regarding Torque 2 in the very near future in this forum area. We are doing our best to get the infrastructure design finalized, and then implemented as soon as possible, and when it is ready we will certainly let everyone know.
That being said, Torque 2 has already completed the primary architecture re-factor iteration which focused on abstracting most of the major systems away from the core engine, such as lighting, GFX (was already mostly complete from the TGE-A code base contribution), and SFX (yes, we have a sound layer abstraction now). In addition, one of the primary areas that has been receiving focus in the early iterations is the separation between engine, application, and tools--in other words, you can look at the TorqueX model, where the engine could be simply used as a set of core modules, the application (your game) interacts with this core module set, and external tools communication work with the application to manage your data.
There are currently 5 full time engineers working on the core architecture and framework iterations, with additional consultants providing support of the development in specific areas.
Once the core Torque 2 engine and framework is stable enough for moving forward, we have an entire (newly formed) business unit of an additional 6 people who's entire focus will be providing a market focused product around Torque 2--specifically reference and user documentation, tutorials and other learning documentation, demos, examples, and other related product units. For historical reference, this side of the development of a product was never before formalized within GG--effectively, we took an engine and pushed it to market, and then developed supporting documentation, tutorials, and demos after the fact, based only on demand. Our entire product lifecycle process is completely different, since we now have the resources as a company (due to the IAC partnership) to take an engine technology and wrap it in a product appropriate for our chosen markets.
Two misconceptions/misunderstandings:
--I made the comment during the roundtable you watched where you got this "2 year" thought, so I'd like to clear it up a bit :) What I wanted to communicate was "if you have a project that has to ship in the next year, Torque 2 is not a primary option. You will not be able to start effectively using Torque 2 in anything but an early adopter phase until sometime in 2008, and since most games will take 3-12 months in the expected markets we serve (and some many more), you won't see games produced by the market using Torque 2 for at least 2 years.
--"we might charge, for top level docs": no change at all here--you are currently charged for books, boot camps (provided for us), or TorqueSchool courses, and this is what Matt was speaking about. We certainly reserve the right to provide additional services around Torque 2, but the "comes with the package" documentation will not only continue to be part of the base price, but will be very much extended in scope, depth, and timeliness. To backup this particular point, I can point to the fact that in the last month, not only have we made base reference documentation for TGE 1.5.2 available to the general public at no charge, but we've also opened TDN to the public as well.
In all honesty, you can't look at historical methods of engine development within GG as an example for moving foward--it would be like thinking about world wide transportation before airplanes were invented--simply not having information about a critical shift in how things are. The IAC partnership is a fundamental change in how we have to do business from day to day--we can now focus on doing things right, instead of "right now" with regards to product preparation and release.
Also a misconception, based on past history (and I'm not faulting you for that at all, it's all you have to go on right now!). None of the products we currently have are "dead"--in fact, if you are planning on publishing a game any time in the next 12-20 months, the existing products are your primary options, and will continue to be the most appropriate. As always, you'll want to select the engine option based on your game type, and most importantly target hardware market.
Torque "2" is exactly that--a fundamentally new engine architecture that builds upon, and corrects where necessary, all of our existing engines. Eventually of course (in the future, in now way at this time), Torque 2 will become the product set and engine of choice for new games, but to use an another analogy, if you need to buy a car in the next 6 months, you won't want to wait until the 2010 models come out--you will use the existing automobile technology that is available today, and make a selection appropriate for your needs.
Once the new transparent development community infrastructure is in place, we'll have quite a bit of additional planning/expectation management information for you (in the coming short months). "idea and no real design" is also completely inaccurate--in fact, in the true spirit of Agile development and "living designs", TorqueX could be considered a design document for Torque 2--the architecture and underlying designs will be very similar. That's not to say that TorqueX isn't a completely viable product (and will continue to be one--in fact the 3D systems and interface is do out here in the very near future), but if you want to peek at some of the inner designs and plans for Torque 2 before the transparent development community is up and running, TorqueX is a very good place to start.
10/11/2007 (10:51 am)
Can't spend as much time as I'd like on this post this morning (we're still at IGC), but wanted to answer a few questions specifically:Quote:
From what I understand (I can be really wrong on this) they only have a design document of Torque 2.
No offense intended, but you are :) Torque 2 is already under active development (and has been for a few months now), and most importantly is using an iterative development process based off of our internal Juggernaut code base, moving forward to the new architecture (modules + components).
Quote:
Oh, first thing concerning "Transparent development" : What is the current state of the project and how much man-power will be dedicated to Torque 2? :)
We are going to be using different web infrastructure from the official GG.com forums/web page for the transparent development program, so I'm not going to be kicking out tons of data regarding Torque 2 in the very near future in this forum area. We are doing our best to get the infrastructure design finalized, and then implemented as soon as possible, and when it is ready we will certainly let everyone know.
That being said, Torque 2 has already completed the primary architecture re-factor iteration which focused on abstracting most of the major systems away from the core engine, such as lighting, GFX (was already mostly complete from the TGE-A code base contribution), and SFX (yes, we have a sound layer abstraction now). In addition, one of the primary areas that has been receiving focus in the early iterations is the separation between engine, application, and tools--in other words, you can look at the TorqueX model, where the engine could be simply used as a set of core modules, the application (your game) interacts with this core module set, and external tools communication work with the application to manage your data.
There are currently 5 full time engineers working on the core architecture and framework iterations, with additional consultants providing support of the development in specific areas.
Once the core Torque 2 engine and framework is stable enough for moving forward, we have an entire (newly formed) business unit of an additional 6 people who's entire focus will be providing a market focused product around Torque 2--specifically reference and user documentation, tutorials and other learning documentation, demos, examples, and other related product units. For historical reference, this side of the development of a product was never before formalized within GG--effectively, we took an engine and pushed it to market, and then developed supporting documentation, tutorials, and demos after the fact, based only on demand. Our entire product lifecycle process is completely different, since we now have the resources as a company (due to the IAC partnership) to take an engine technology and wrap it in a product appropriate for our chosen markets.
Quote:
Wont be for at least 2 years as I understood from the stream I watched. And I sure hope that the coming docs will be better than the TDN, but I did a lil jump when they mentioned that 'we might charge, for top level docs'.
Two misconceptions/misunderstandings:
--I made the comment during the roundtable you watched where you got this "2 year" thought, so I'd like to clear it up a bit :) What I wanted to communicate was "if you have a project that has to ship in the next year, Torque 2 is not a primary option. You will not be able to start effectively using Torque 2 in anything but an early adopter phase until sometime in 2008, and since most games will take 3-12 months in the expected markets we serve (and some many more), you won't see games produced by the market using Torque 2 for at least 2 years.
--"we might charge, for top level docs": no change at all here--you are currently charged for books, boot camps (provided for us), or TorqueSchool courses, and this is what Matt was speaking about. We certainly reserve the right to provide additional services around Torque 2, but the "comes with the package" documentation will not only continue to be part of the base price, but will be very much extended in scope, depth, and timeliness. To backup this particular point, I can point to the fact that in the last month, not only have we made base reference documentation for TGE 1.5.2 available to the general public at no charge, but we've also opened TDN to the public as well.
Quote:
It all sounded truly fantastic, but WAY out of scope for GG if history has taught us anything as we have waited and continue to wait on a truly release version of TGEA.
In all honesty, you can't look at historical methods of engine development within GG as an example for moving foward--it would be like thinking about world wide transportation before airplanes were invented--simply not having information about a critical shift in how things are. The IAC partnership is a fundamental change in how we have to do business from day to day--we can now focus on doing things right, instead of "right now" with regards to product preparation and release.
Quote:
The basics I took away from the future discussions is that TGE is dead, TGEA will be the "tge", and Torque2 will be the "new" TGEA (relatively speaking).
Also a misconception, based on past history (and I'm not faulting you for that at all, it's all you have to go on right now!). None of the products we currently have are "dead"--in fact, if you are planning on publishing a game any time in the next 12-20 months, the existing products are your primary options, and will continue to be the most appropriate. As always, you'll want to select the engine option based on your game type, and most importantly target hardware market.
Torque "2" is exactly that--a fundamentally new engine architecture that builds upon, and corrects where necessary, all of our existing engines. Eventually of course (in the future, in now way at this time), Torque 2 will become the product set and engine of choice for new games, but to use an another analogy, if you need to buy a car in the next 6 months, you won't want to wait until the 2010 models come out--you will use the existing automobile technology that is available today, and make a selection appropriate for your needs.
Quote:
No real timelines were discussed, and I also got the impression that all they have at this point is an idea and no real design at this point.
Once the new transparent development community infrastructure is in place, we'll have quite a bit of additional planning/expectation management information for you (in the coming short months). "idea and no real design" is also completely inaccurate--in fact, in the true spirit of Agile development and "living designs", TorqueX could be considered a design document for Torque 2--the architecture and underlying designs will be very similar. That's not to say that TorqueX isn't a completely viable product (and will continue to be one--in fact the 3D systems and interface is do out here in the very near future), but if you want to peek at some of the inner designs and plans for Torque 2 before the transparent development community is up and running, TorqueX is a very good place to start.
#9
Over the last 12-15 months, our development process has gone from being very open (communicating what we were working on, priorities, etc.), to being 100% opaque--we haven't been talking about products until after they have already shipped. Quite honestly, we've disliked this approach even more than you have--and in the last 4 months we've had a half dozen very smart people brainstorming and evaluating how to reverse this trend--and "transparent development" is exactly what we mean--you, as a community, will be able to observe and discuss, and eventually interact with the ongoing development of a new engine.
Since it's such a fundamental change from how things have been over the last year, it's really important that you as the community allow this to sink in, and treat it appropriately. We listened to your concerns and frustrations with opaque development, and responded--our next generation engine is going to be developed before your eyes, with as much communication about design, progress, iterations, "ready" capabilities, and other information as possible, as the engine is developed.
There was a post above regarding a mostly negative connotation of "early adopter"--for valid reasons. This new shift to transparent development honestly needs to be considered as a totally new shift from "early adopter"--and I realize you have nothing to base your trust on the statement I just made--all the previous experience has indicated otherwise. The raw fact of the matter however is that we're dead serious about this--and we ask that you work within the transparent development process in a mature manner, and respect it for what it will be, and that you take our actions as well as our words into account when evaluating the transparent development process.
I fully respect that it's going to be difficult to break from what we've shown you in the past, and treat the intentions behind transparent development for what they really are--allowing our community to observe the development process. Quite frankly, were I on the other side of the Internet right now, reading this message, I'd think "yeah, sure--it's going to be early adopter TSE all over again"--all I can ask is that you give us both the time to demonstrate our intent, and the maturity and understanding required to allow a transparent development process to work, and evaluate the results at the end of, and of course during, the process, instead of now when it's barely been announced for 48 hours :)
10/11/2007 (11:04 am)
One final thought here that is fundamentally important for everyone to understand:Over the last 12-15 months, our development process has gone from being very open (communicating what we were working on, priorities, etc.), to being 100% opaque--we haven't been talking about products until after they have already shipped. Quite honestly, we've disliked this approach even more than you have--and in the last 4 months we've had a half dozen very smart people brainstorming and evaluating how to reverse this trend--and "transparent development" is exactly what we mean--you, as a community, will be able to observe and discuss, and eventually interact with the ongoing development of a new engine.
Since it's such a fundamental change from how things have been over the last year, it's really important that you as the community allow this to sink in, and treat it appropriately. We listened to your concerns and frustrations with opaque development, and responded--our next generation engine is going to be developed before your eyes, with as much communication about design, progress, iterations, "ready" capabilities, and other information as possible, as the engine is developed.
There was a post above regarding a mostly negative connotation of "early adopter"--for valid reasons. This new shift to transparent development honestly needs to be considered as a totally new shift from "early adopter"--and I realize you have nothing to base your trust on the statement I just made--all the previous experience has indicated otherwise. The raw fact of the matter however is that we're dead serious about this--and we ask that you work within the transparent development process in a mature manner, and respect it for what it will be, and that you take our actions as well as our words into account when evaluating the transparent development process.
I fully respect that it's going to be difficult to break from what we've shown you in the past, and treat the intentions behind transparent development for what they really are--allowing our community to observe the development process. Quite frankly, were I on the other side of the Internet right now, reading this message, I'd think "yeah, sure--it's going to be early adopter TSE all over again"--all I can ask is that you give us both the time to demonstrate our intent, and the maturity and understanding required to allow a transparent development process to work, and evaluate the results at the end of, and of course during, the process, instead of now when it's barely been announced for 48 hours :)
#10
I suppose we have to expect a break in the legacy format, and everything will have to be reworked from scratch - my concern is here on the artist side, not coding?
10/11/2007 (11:09 am)
One question!I suppose we have to expect a break in the legacy format, and everything will have to be reworked from scratch - my concern is here on the artist side, not coding?
#11
--Collada is, moving forward, going to be one of the model formats of choice. If your modeling application supports that format, you'll be fine.
--legacy support of DTS is most likely a no-brainer--as well as importers for things like existing mission files, etc. From an asset perspective, things will still remain relatively legacy supportive--the fundamental changes are going to be in game design (more information on components and games built around an aggregation model will be coming in the future). To set an expectation early: there is no expectation of backwards compatibility with games as a whole from existing product lines--you aren't going to be able to take a full TGB, TGE, or TGE-A game and suddenly "port" it to Torque 2 without quite a bit of work: Torque 2 is being designed as a new engine, and intended for new projects. That being said, it does make sense to us to provide as much flexibility as possible for existing assets.
One more add: Also mentioned in Clark's presentation was that we will be moving to a poly soup rendering/collision model as our primary internal representation. However, given the modular architecture that Torque 2 supports, physics, rendering (as well as lighting, input, networking, and a host of other modules) are designed to be "drop in and compile"--including external/3rd party implementations if you so choose. The core work going on right now at the lowest levels of the engine architecture are built around the concept of full modularity--removing dependencies and directly coupled tied between systems towards an "exposed interface" scenario where changing a specific module has little to no effect on other selected modules within your application.
10/11/2007 (11:15 am)
To early to do much more than speculate to be honest, but some key points from Clark's session:--Collada is, moving forward, going to be one of the model formats of choice. If your modeling application supports that format, you'll be fine.
--legacy support of DTS is most likely a no-brainer--as well as importers for things like existing mission files, etc. From an asset perspective, things will still remain relatively legacy supportive--the fundamental changes are going to be in game design (more information on components and games built around an aggregation model will be coming in the future). To set an expectation early: there is no expectation of backwards compatibility with games as a whole from existing product lines--you aren't going to be able to take a full TGB, TGE, or TGE-A game and suddenly "port" it to Torque 2 without quite a bit of work: Torque 2 is being designed as a new engine, and intended for new projects. That being said, it does make sense to us to provide as much flexibility as possible for existing assets.
One more add: Also mentioned in Clark's presentation was that we will be moving to a poly soup rendering/collision model as our primary internal representation. However, given the modular architecture that Torque 2 supports, physics, rendering (as well as lighting, input, networking, and a host of other modules) are designed to be "drop in and compile"--including external/3rd party implementations if you so choose. The core work going on right now at the lowest levels of the engine architecture are built around the concept of full modularity--removing dependencies and directly coupled tied between systems towards an "exposed interface" scenario where changing a specific module has little to no effect on other selected modules within your application.
#13
Gary (-;
10/11/2007 (11:21 am)
Stephen; how do Torque2 and IA work together? Will Torque2 be able to be used as the engine for these browser-based games?Gary (-;
#14
10/11/2007 (11:21 am)
Quick note: I'm late for a meeting at IGC, so will not be back to answer further questions for several hours--please be patient :)
#15
10/11/2007 (11:23 am)
Not my area of focus so I can't say a ton, but InstantAction itself is technology neutral. Torque engines (existing and future) will of course have a "leg up" in using IA.com, but it's not required in any way.
#16
10/11/2007 (11:23 am)
One of the things we've been going back and forth on irc a bit about at the moment was the component system and behaviors, wich brings up a couple points, actually: how much live discussion should we expect to participate in as a comunity, and just how deeply bifurcated is the component/behaviors system looking to be, so we don't end up with yet another shapebase situation embeded someplace?
#17
Gary (-;
10/11/2007 (11:30 am)
How do games play in a browser, but be technology-neutral? Obviously C++ games need to work in it?Gary (-;
#18
10/11/2007 (11:37 am)
Woah, thanks a lot Stephen for your openess and clarifying the rumours about Torque 2 that have been floating around for some time already. For me it seems the first "influences" from the IAC deal are already visible with this news about Torque 2 and I have to say I like it. :-)
#19
10/11/2007 (11:50 am)
Sounds great...My project is leaning towards polysoup and i will welcome this change :)
#20
I just have question concerning poly soup. From an artists point of view I really like the idea because of the obvious freedoms it would provide. But I'm also concerned that for large scale games with large environments that poly-soup will not be a truly viable option. So my question is, that if my concern is correct and there still will be a need for BSP is GG going to provide that module for the T2 core? Or would that be something that developers and the community will have to deal with?
Thank you again,
James D,
10/11/2007 (11:57 am)
First I'd like to thank Stephen Zepp for taking the time to answer questions and remove doubt so soon after T2 was announced. I just have question concerning poly soup. From an artists point of view I really like the idea because of the obvious freedoms it would provide. But I'm also concerned that for large scale games with large environments that poly-soup will not be a truly viable option. So my question is, that if my concern is correct and there still will be a need for BSP is GG going to provide that module for the T2 core? Or would that be something that developers and the community will have to deal with?
Thank you again,
James D,
Torque Owner Mathieu
From what I understand (I can be really wrong on this) they only have a design document of Torque 2.
Oh, first thing concerning "Transparent development" : What is the current state of the project and how much man-power will be dedicated to Torque 2? :)