Torque game engine

Started by
8 comments, last by BobV 19 years ago
Some friends and I were looking into using the Torque engine for a project between ourselves. The price ($100) I believe is well within our capability, and it looks pretty powerful. However, I wanted to check here first and get some opinions on it. Has anyone here used it? Any opinions on power/speed/usefulness? It seems rather nice, but I've never used it or any other premade engine like that. Thanks for any help!
Advertisement
Check out Cloudburst in the GDS section for a nice display of the Torque2D engine.

Jason
Quote:Original post by Malderi
However, I wanted to check here first and get some opinions on it. Has anyone here used it? Any opinions on power/speed/usefulness? It seems rather nice, but I've never used it or any other premade engine like that.


It's powerful, fast, and useful. How's that :)? Seriously, as with anything else you have to know what you are doing to get the most out of it. It's not for the faint of heart. A good way to get an idea of what Torque can do is to download the Torque Demo, along with demos of some of the games for sale at GarageGames (not all of them are Torque games, but several are - off the top of my head: Marble Blast, Orbz, Think Tanks, and Dark Horizons: Lore are all Torque games - and there are others). There's nothing like production code that will tell you what an engine can do.

If you do go with Torque, I also suggest that you pick up a copy of Ken Finney's excellent book 3D Game Programming All In One. It will help you get up to speed with Torque much faster than you would without it (the scripting side anyway, which is half the battle). Even so, you should expect that it will take some time to learn your way around the engine well enough to make it sing, as it's a complex beast. It's not something you are going to pick up in a day or a week. Also keep in mind that it's $100 per seat to license Torque, so if you have 3 programmers on your team that's $300 if each of them will be touching Torque.
Yeah, I realized that part of it. My team is kind of deciding right now whether we want to go with Torque, or pick up another graphics library like OpenSceneGraph to use instead. We could probably program directly from OpenGL or DirectX, but we'd rather spend our dev time on other things since it's a small team.

What I'm really looking for, though, is any first-hand stories from programmers using it. You sound like you have. Any good/bad/ugly stories to relate? Parts of the code that are just damn freaking cool, and others that you wouldn't touch with a ten-foot pole?
Every story I've heard about torque went something like:
"Its a nice and powerful engine as long as you stick to using the scripting engine, but once you try to modify the actual engine source, it is a nightmare."
I've no idea how true that is, or how powerful the scripting system is.
"Walk not the trodden path, for it has borne it's burden." -John, Flying Monk
I don't know what your coding level is, but I would also suggest the OGRE engine. It's free, and that's a plus. It doesn't have a built in scripting engine as it's a rendering engine, not a full on game engine. It is really easy to work with as far as building your own stuff. If the problem the above poster mentioned is an issue( not being able to work outside the scripting engine ) then take a look at OGRE and see if it would work for you.

karg
Making a Torque game can seem more like making a mod than a truly original game, because they encourage you to stick with the scripting and avoid code changes if you can. True you do get source code, a LOT of source code. Somewhere around 500,000 lines. I found that I got lost in the sheer size of the code, plus you need to struggle with the complexities of the scripting system too.
Value of good ideas: 10 cents per dozen.Implementation of the good ideas: Priceless.Machines, Anarchy and Destruction - A 3D action sim with a hint of strategy
if you want to use shaders (specified by CgFX) very intensively i would recommend having a look at nVidia's Scenegraph "NVSG" @

http://developer.nvidia.com/object/nvsg_home.html

It has Linux and Windows support even for 64 bit processors.


In conjunction with the lua scripting language

www.lua.org

it would be my recommendation.

cheers.
Quote:Original post by Malderi
What I'm really looking for, though, is any first-hand stories from programmers using it. You sound like you have. Any good/bad/ugly stories to relate? Parts of the code that are just damn freaking cool, and others that you wouldn't touch with a ten-foot pole?


I don't know how to best answer this, so I'll try to generalize. For several reasons, the architecture of TGE (Torque Game Engine) is not as modular as one would like. Several modules are tightly coupled. Some areas of the code are well commented, others are not, and still others leave you scratching your head. Any significant project will require modification of the C++ source, no matter how much they encourage you to use TorqueScript. You will almost certainly want features that aren't built into the stock codebase. The plus side to that is that many others have been down that road before you and have posted resources on the GG community site, which you can download and apply. The downside is that some of those resources are old and require tweaking to work with the latest release. In general, adding new features is not a big deal, but sometimes it can take significant effort - particularly when no resource exists to get you started.

The worst part of the engine, IMO, is the terrain system. It's fast, it's pretty, and it works. But it doesn't fit all projects and is quite limited in the amount of flexibility you have with it. An ongoing discussion in the community has revolved around revamping the terrain engine to support paging (currently, it infinitely repeats, but a resouce exists showing you how to make it *not* repeat) and to give it more flexibility in general. Code-wise, it's a mess. The terrain engine relies so heavily on some texture blending code which is optimized for specific parameters, that tweaking terrain parameters is out of the question - and rewriting the blender is a herculean task. Plus, the terrain system is so tightly integrated with the rest of the engine that replacing it entirely will require rewriting/replacing/removing huge chunks of code from different modules. There were a couple of projects that got off the ground in that direction, but neither is active any longer, nor compatible with the current codebase. However, one guy recently developed and released a terrain deformation resource (no small feat considering the networking architecture coupled with the aforementioned difficulties), and another fellow is working on what he calls an MMO pack which will include a terrain paging system for TGE (among other things). The TSE (Torque Shader Engine) is much more modular, from what I understand, and includes a much, much, improved terrain engine that has no limitations on parameter tweaking or extending. Also, the next version of TGE (1.4) will, supposedly, be more modular and have some improvements to existing modules. To what extent, and if it will affect the terrain engine, is anybody's guess.

Take a look at the Torque2D engine. That is one freaking cool mod to stock TGE. But Melv May, the guy who created it, didn't get into it overnight. He spent several months learning Torque inside and out, releasing resources as he went (several of which got integrated into the stock codebase). And he's an experienced programmer. He was at it for sometime before he was able to get to a point where he could create T2D.

My opinion comes after having done quite a bit of playing around with TGE. Four years ago I would have said flat out to stay away from it, it's a beast! But I've learned enough about the architecture now that I have an idea where to start when I want to make modifications. One of the things I found helpful was to strip the engine down to the bare minimum - i.e. the core platform code and script engine. Going from that, building it back up again by adding modules in piece-by-piece really gave me a lot of insight into what was going on. So what I'm getting at is that, even if you are an experienced C++ programmer, it will take a fair amount of time to arrive at a level where you can easily make modifications - and even then some modifications will require more work and/or a greater challenge than you expect. That's why I find your question difficult to answer. For a first time TGE user, it is extremely challenging to work with. Given time, it gets easier, but how much time is up to you.
Just to throw in a possible alternative. I've recently been looking at the delta3d engine at www.delta3d.org (open source). I've been rather impressed by the way they've setup everything from install to documentation. I haven't been using it long enough to give it a solid thumbs up or be able to compare it to other engines but the graphics side is built on top of open scene graph. The team seems to be very productive and from what I gathered from the website, I think they are fully funded and working on this full time. I'm hopeful that we'll see great things from this engine. Might be worth checking out before you commit to something you need to pay for.

This topic is closed to new replies.

Advertisement