Who wasted there money? :P

#81 I started with TGE in '02. I'd held off because I didn't know or have VC++. One day saw it on sale at the local office supply store. I took it home, installed it and a couple of hours with the online docs had compiled the "Hello World" App. With my esteem thus bolstered, I downloaded TGE and spent the evening going through the steps on this site. According to my backup CD from that night, I had FPS running six hours after downloading the code. This, after having never compiled a program 'til that afternoon.

I was not then, nor am I now, a 'programmer' but there was sufficient documentation even back then for a total n00b to get it working. Granted there are things I wish were different (audio being one of them), and I have had times of frustration, but all in all it is well worth the money and time I've spent. FWIW this is still one of the best online gamedev communities, and any shortcomings (IMHO they are very few and very minor) the engine may have are more than made up for by the willingness of folks here to help out...

#82 Yes I agree that Torque can't be the perfect solution to every project, and that's a valid point you bring up Stephen.

Though I would like to add that I did look at starting from the beginning and it was quite an undertaking. I looked at deriving from ShapeBase, but then found that it also had fps integrated into it. So then I looked at deriving from GameBase and found that it might work. But in short, trying to figure out all the systems just to make a simple application was tough (of course this was when Torque was lightly documented, I look through the code now and it's much better).

Yes going to some other system you have to do the same things. The only difference is that other systems usually teach you from the beginning how the system works. All of the Torque samples were based off the fps stuff, so they didn't help much if you wanted to do something else.

It's kinda like learning a programming language. If you presented a person new to programming with a 500 line example of how to use the language, the person would have trouble understanding. But if you show them how to use an if statement, then a for loop, it makes it much easier to follow.

So for Torque it would have been nice to see:
"minimal application" - (I actually found this in the resources now!)
"console" - The minimum it takes to get console running
"gui" - The minimum to load a gui
"input" - The minimum to handle input
...
etc.

And just for example. It took me weeks to figure out the terrain handling in Torque, and I still had trouble with it. While in openscenegraph I had a terrain up and running in a day.

Anyways I Hope this could be of some use, you all do good work. Thanks for your dedication to indie developers.

#83 Some concepts only make sense when you've got 500 lines, though...

ShapeBase is a pretty good basis for any sort of game. Is it overpowered for almost everything? Yes. Does it try to do too much? Yes. But you can definitely build a lot on it, fast. The RTS pack uses ShapeBase! Puzzle games use ShapeBase! Orbz uses ShapeBase! Phil's flying game does! A game I played once with penguins did!

Does it have obvious problems? Is it gonna get replaced with something better? Absolutely!

But it's also REALLY useful. It wouldn't be tolerated if it wasn't, and to throw it out sight unseen is a bit silly, in my opinion...

I'm not sure why you had so much trouble with terrain. I was able to get a basic terrain up and running (for some shader testing in TSE last year) in a few hours. It was gross and ugly, sure, but it also worked great for testing some light scattering techniques on. The terrain code in TGE is a bit gross and scary, but it's only a matter of hours to drop in a new object and use it instead. If you want to rewrite the terrain anyway, why not do that? :)

#84 I almost gave up as well in the beginning when I could not easily set up a custom RTS style camera. I found it easier to set up a flying vehicle as a camera platform in TGE. Coming from something like Blitz3D where it would be simple to do this was very frustrating for me. I had accomplished a lot in a short time with Blitz3D and was a bit spoiled. I had completed an RTS style demo with mouse picking, marquee selection, waypoints, terrain, animated character, and custom camera very easily. I also worked on a tiling deformable 3d terrain system and model viewer.

Some old images of that work:

Of course what I learned is that I would much rather take the time to learn a tool like TGE so I would not have to build everything from scratch and could spend more time on art content and design. 3D Game Programming All In One got me back into TGE and was exactly what I needed. I know that Torque is a great application, but it can be very frustrating for the new user. I think the new user purchases TGE and is thrown into the middle of a stripped down game then left stranded with way too many things to absorb at once.

There is something to be said about starting with a blank canvas. When I first got TGE there were no MinApp Tutorials. Thanks to James Yong and John Vanderbeck we have them now. I just wish GG had put something like this together to begin with. Seriously, without people in the Torque community putting together a book, tutorials, and add-ons I probably would have given up on TGE a long time ago.

Another frustration for me has always been the Art Content Pipeline using Map2Dif, QuArK, and Milkshape3D. I worked with David Mathews to get support for Dif and Dts into Greenbriar Studio's Toolbox to improve the process. We eventually got static Dts files to work, along with some animated Dts support, but Difs were just not working right even though they loaded fine into Hammer and QuArK. Somewhere along the way David's GG contact stopped replying to his emails and TGE support in Toolbox got shelved.

However, I've watched the Art Content Pipeline mature a little with the addition of Show Tool Pro and the announcement of Constructor and am getting excited about TGE/TSE again. I'm looking forward to maturing alongside this product and have learned a lot using it. Hopefully I won't get too spoiled by Virtools over the next 3 semesters though. Once I am out of school am hoping to use TGE/TSE professionally and it looks like the timing will be perfect.

#85 Well the terrain I tried messing with back in 2002, and I was just trying to change the current one (yes it was a mess and scary), and I was trying to build a clean app with only the terrain and without all the extra fluff.

Quote:There is something to be said about starting with a blank canvas.

Well said.

#86 I've been watching this thread and it's interesting to see peoples view points. I hadn't planned to reply but I do want to reply to Ben Garneys statement.

Ben, your example of what you did with the terrain is interesting but I think you lost sight of where you are with the product versus your customers. With the exception of those in school, your indie developing customers are only able to work on their projects for a few hours at a clip. This is a much different programming dynamic compared to the situation you're in. You work with the product full time, you developed the terrain code, and you work with those who have developed Torque along side you. Using your experience as an example is unfair in that you are closer to the product then any of us can ever hope.

Indie development doesn't pay the bills and for those of us with families and children weekends and holidays are not in the cards. So try to imagine what you did without being able to talk to the guy next to you and at three to four hours at a clip. By the time you get into the development groove it's almost bed time so your fresh for your 9-5 job.

I'm not blaming or attacking you, I'm in the same position. I am developing our product from scratch and I'm not qualified to work on documentation. I'm too close and make many assumptions no matter how hard I try. My users tell me I'm out of touch and that my answers appear like magic while to me they appear obvious. :) You and I are in the same position and as architects we have to remind ourselves that our customers do not have the same advantages.

#87 I'm sorry Joe, although I see where you're coming from, I also agree with Ben.

I also work a full time job. I also have to squeeze Torque in after work or on weekends. I don't have kids so I guess I'm lucky, but still. I don't get to work with Torque all day either. I don't have a Torque expert sitting next to me when I'm deving (Well, I have the forums).

It doesn't matter.

No matter what engine I use, whether I use Torque or start from scratch, I will have to work under the same time frame. And the fact remains, to build an engine even vaguely comparable to Torque would take far longer than figuring Torque out.

Game development is hard, and takes a long time. A hellova long time. The pros take years, even though they're working full time. You gotta accept that squeezing in game dev after hours is gonna be incredibly tough, and just live with it. Torque helps, but you've still chosen to climb Everest. Its gonna be challenging. GG is doing an amazing job, but they can't perform miracles. Anyone working under those conditions you mentioned needs to have herculean endurance (and not be overambitious ;-) ).

I'm sorry, but I've just seen too many people lash out at Torque for not being easy for "people who only have a few hours" to learn. But they're assuming that its possible to make this task "easy". Its not, not with a commercial class engine. Especially for people new to game programming concepts. It took an entire team of people years and a huge budget to make Torque, but you get people coming online and complaining that they've spent "hours" and still don't really get how torque works. >:-|

I dunno, maybe after spending 6 months working on my own engine, and getting as far as a nice terrain renderer with some trees, I'm more appreciative of what Torque offers

#88 I can somewhat agree with both of you... While yes, Ben is very close to the code and knows it like noone else (or few people)... It doesn't mean that other people cant learn the code and do amazing things with the engine. I have a 6:30 - 3 job and I spend 7 hours each night working on learning and utilizing the engine. pepole call me nuts, but the more dedicated you are the easier it becomes to find the time.

You'll never just know Torque.. Im sure even Ben has to read the code and read the docs a lot of the time... Perhaps instead of learning the engine, we should be trying to leanr the root basics, and learning how to better utilize the documentation and resources available to us...

Like C++... I don't know most of its libraries... but I know how to use my trusty reference manual and resources on the net... So I can do ANYTHING with C++, and as i learn to use my resources better, the faster I become.

With a little reserach and some planning you could rewrite the terrain... You just have to put the job into a feasible scope.

#89 Gareth,

I think you and I are in agreement. I'm not discounting what Torque accomplishes and I'm not turinng my license in or defuncting the product any time soon.

Your right in that it's worth the $100.00. In fact I think it's too cheap but that's another topic for another day.

I'm only taking exception with Bens statement "I'm not sure why you had so much trouble with terrain. I was able to get a basic terrain up and running (for some shader testing in TSE last year) in a few hours."

I probably should have made that more clear.

By the way I do have to take exception to one comment you made, "I don't have kids so I guess I'm lucky". If you had kids you wouldn't say that. :) Kids are the best and most difficult Game Engine. :)

#90 Hi Joe.

Let me qualify my statement a litttle bit. I took the fxRenderObject, and then I put a draw call in it to draw a vertex buffer that contained a very simple, precalcuated mesh. This was not a full-on terrain implementation but rather a simple placeholder object to test some things with. It was literally less than fifty lines of code added to fxRenderObject. :)

This was on TSE, with a very rough version of the GFX layer in place. To do the same thing in TGE, with a working, reliable graphics layer (ie, OpenGL) shared with many, many other applications, would be even easier.

I understand your concern, and I realize that I'm very close to the engine. That's why I chose the example I did - it's something that anyone with a familiarity with general 3d graphics programming, after a few weeks of "spare time" research on Torque-specific issues, could do. If you had a working terrain engine that was written on OGL, I would guess that a "full time Torque programmer" like myself would be able to port it in a few weeks, while for someone who had to learn as they went it might take several months. If you have to have the tech, this is not a high cost to pay, and I think well within reason for a spare time project.

This all assumes that you have a basic understanding of the terrain code you're porting, general 3d graphics concepts (transforms, textures, render states, etc.), and experience with at least one major graphics API (OGL, D3D, even GLIDE). If you're trying to do serious development on a 3d game, I think these are pretty reasonable! Maybe my assumptions are flawed. :)

#91 I think your assumptions are correct.

Quote:after a few weeks of "spare time" research on Torque-specific issues

That's my point. I had to research for a few weeks to figure out the terrains working with the engine. In OpenSceneGraph I looked at a terrain example and plopped in my custom terrain (I do have more experience now so that may be partly to blame).

NOTE:This is nothing against Torque and is not a Torque vs OpenSceneGraph comparison and not a rant.

I'm just merely trying to make the point that starting from scratch at least has some value and makes things at least seem simpler as well as letting you get off the ground quicker. I believe it really is simpler.

The coding example is still the best that I can think of. Think of it like this:

If I know how to program and I wanted to learn a new language, it would be easier to look at and understand in pieces instead of a finished product that used the pieces. For example when I learned Python, at the time I knew quite a few languages (c++, asm, fortran, basic, etc.). Looking at a Python program hardly helped me at all. But when it was broken down to

Here is a list:
li = ["item1", "item2"]
and here is how to get the last item in a list:
last = li[-1]

I then exclaimed, Oh! I get it now. I had already used lists in c++ and other languages, I knew how they worked. But I wasn't familiar with how Python used lists.

The same I believe is true with Torque. There were lots of scripting examples to show how that was done (and it's much simpler to mess with scripts). But very few (if any) examples as to how the engine could be used (note, this was in the past).

#92 This actually underscores a major difference between Torque and OpenSceneGraph. OSG is meant to be a vehicle for rendering technologies. Torque is meant to be a vehicle for game development. Thus, OSG comes with lots of examples for doing terrain rendering, setting up different sorts of scene objects, and so on. Torque, on the other hand, comes with lots of examples for doing game-like tasks - vehicles, players, interesting uses of collision primitives, scriptable objects, and so forth.

We are working on several initiatives to make it a lot easier to do engine changes, though it will probably be a long time, if ever, before you'll see docs like "how to replace Torque's terrain engine with another one that works just as well." We already have the "how to make an object that renders itself" doc, which I myself refer to frequently and find incredibly useful.

#93 I bought the TGE in December 04, so just a few mths ago, it is collecting dust as I am so lost with it, learning to use modwizard got me wanting to do more, I have less of an idea how to use torque then most of the ppl who buy it. It is not a waiste of money even if it sits on a shelf. I will learn to make games in due time as I learn. and when this happens I will have one of the best tools in this engine, as this is common knowledge. But if you think it is a waiste of money it's only because of your lack of knowledge of useing it. So do as I am going to do, step it up and learn how to use this engine and stop whinning.. lol

#94 I have spurts of development ... one month I will be right into it and then maybe for a couple of months no time or urge to develop/play with something ...

I bought the RTS and haven't touched it yet, so I guess, that's a yes from me, but I intend on using it, so I wouldn't say it was a waste of money.

#95 Like anything: If you don't use it then it's a waste of money for sure... Here's a small story to illustrate the issue:

About 7 years ago I bought a bike that was intended to be used as incentive to
a) quit smoking
b) lose some weight...

etc, health issues etc..

So that bike sat in various places around my house entirely unused for about 2.5 years... Then one day I said to hell with this and quit smokeing...

That bike went from 0 use / waste of money to being a critical part of my life... To date I've replaced it 3 times (1 mechanical failure, 1 accident) Current incarnation is makeing me pretty happy...

Same thing with Torque... It could sit there for ages and then your situation changes and it could be come important...

For myself, I bought it for the purposes of makeing demo's for potential employers...

I am haveing *some* issues in that regard because most employers don't want to see a ton of scripts... So, the "meat" of anything I do has to be in C++... There's some lack of proper docs for that purpose...

#96 What holes are you finding in the docs?

#97 I haven't put Torque down since our company bought it. Our greatest concern was with Torque's performance, in that we weren't sure how robust the engine was going to be. We shouldn't have worried....

I, personally, am an experienced DirectX programmer (re: C++) so the scripting aspect of Torque came as a breath of fresh air. Still, having the background of programming in C++ has been a huge benefit. The compiled version of Torque that I work with has several additions to it (Database support, additional graphics routines, cross-platform communication, etc.).

The Torque engine is great for Indies from the standpoint that it can considerably shorten your production schedule. But it won't make you a power programmer if you aren't one already.

Bottom line: $100 is an absolute steal.

#98 I just bought my license and I have enjoyed playing with it. But as previous comments have stated that the learning curve is quite steep.

We use Torque at my school as our primary engine to learn with, I purchased the license cause I wanted to use it at home. If it wasn't for the fact that a group of fellow students and I have been almost quite literally going through each line of code, and trying to figure out what it does and then comment it we would be lost. Its been a long hard struggle but we have a pretty decent internal website with our documentation which suprisingly enough was just supposed to be for my small group but is now used as a teaching tool in the class room.

The bottom line comes to this, if you are dedicated and don't give up this is probably the best deal any game developer can get.

#99 The original poster should have saved his money and used it to buy a spell checker. ;)

I've been thinking about the purchase since it first became available, and finally found an "excuse" to close. I'm now working on a team as an engine coder. My specialties are in debugging and communications code. I've been a Tribes 2 server admin for a couple of years and was always frustrated when it would mysteriously crash or when things would fail with inadequate logging. With an engine source license I can now work on a game as an effective troubleshooter, drilling down into the bowels of the beast to find root cause.

So far I have to report that the code quality is very good, and the only bugs I've found were introduced from the addition of 3rd party resources. (They were easy enough to find by inspection using a gdb backtrace.)

#100

Quote:
I've asked a qualified (and degreed) Software Engineer to evaluate the code base.

But was he also licensed to see the code?