Torque and the FPS Template Debate

#1 The problem a lot of people have is that the code limits what you can do. If the engine is coded to be FPS-centric, no amount of tools will let you create an RTS. It just doesn't happen, at least not without a lot of pain that artsy types won't be willing to fight their way through (and should't have to).

This is a problem with stock templates, but also with the engine source code itself. A better art pipeline is good, but still doesn't solve fundamental issues.

#2 My 2 cents.... I get it, FPS limits code, so does RPG, TPS, RTS etc etc. I suggest STOP thinking 'templates'. If you want to 'sell' the engine, then build an amazing demo or series of demos. Hell, look at what Blender has done. Build a small team, build a 'demo' using that team. Take what was learned and add that into the next 'release' candidate. Build enough good stuff and (as blender proved) you can even get hard ass, old school Max guys to totally convert. Granted, this is NOT an easy approach but, it freakin works.

As for 'templates' (I can't believe I am saying this...) Unity has it right. (I know we were not supposed to mention it) Start with a blank canvas. Let the designer determine what is needed from the get go. (ANNNNDDDD, since we dev the 'demo' levels as MIT, and add in the 'features and tools'. Why not release those sources as community property?)... see where I am goin with this?

Just my 'quick' thoughts. OH YEAH.... and the 'team members' well they get a little something called practical experience deving a REAL game... I know I would be 'up for that'.

It is a solid 'proven' way to develop a product. Yeah, financing can be an issue etc. Blender did it, others have done it, if T3D is going to MOVE forward, I think we need to 'reevaluate' what the 'steering committee' is and means. Let's build something rather than just put out a release every few months with some 'neat' stuff, but NOONE knows how to use or benefit from.

NOT HACKING ON THE SC... You guys are doing great, I just want to see this engine be the 'default' for indie game dev and I think this one way to get there.

Ron

#3
I agree with you Daniel but I still think templates are just specific to a game. coincidentally the current FPS template is just a demo with torque and not the basis or guide of making a game or what torque is all about. Perhaps what must be done is create new functions that is general and will cater to all game play genres, and expose some of that with torquescript. I guess UDK solves that code problem with kismet.

torque source code is already there with API so game dev can later extend the engine's functionality to their specific game requirements. and I have seen some amazing kits (code) here that really looks awesome and better, compare with other 3rd party kits from other engines. ie:

ubiqvisuals adventure kit
recast navigation (again great work btw)
sahara environment
GMK (another great tool)
UAISK (somehow dead but looks good)
air combat kit or sth
RTS and other templates
other kits on store.. so on

built-in tools:
torque road editor, mesh and texture (still unrivaled, unity and udk doesnt have any and in unity you have to buy a mediocre 3rd party separate pack for it http://www.unityterraintools.com/tools.php)

I have not tested all of them but I already read some good feedback . so personally, I don't see any problem with torque "code-wise" if I will be working with a serious indie c++ programmer or script coder. it's the need to *present* torque again to the gaming dev public and new users and make them see what they're missing out.

#4 great points Ron. yes, the current SC is still a great team with great dev experience. I just think the game plan is just not "fine tuned" at the moment.

I don't mind starting from a blank slate, just as long as I have the tools to build a complete and competitive scene to present a demo using the engine.

#5 GamerX,

Agree but, right now, we have an engine that is going in 500 different directions. That is fine however, it is not easy to be great at everything and a master or nothing. Seriously, I have NO idea how the latest build helps me as an environment artist. 64bit, 32bit, I get the same frames per second either build. WTF? Now, if I had the time and experience to work DIRECTLY with the coders and scripters (aka a small team) I could probably optimize my output to match their intent. Hence, my suggestion. Heck I think I even recall a post where someone said 'screw the FPS stuff make an RPG version because that is all anyone wants'. Probably not that harsh but you get the idea. As a former 'member' of the steering committee, I understand the time and effort they have to put in. Hell, just try doing a GIT merge.... good lord. (on top of that, make sure all the code you are merging does not break another aspect of T3D...don't miss that at all).

I understand there are 1000 people all trying to make 'their' game with T3D. I get that each 'genre' has it's own idiosyncratic issues. However, the only way we can move ahead is to work through each 'issue' for each genre one feature at a time. If we do this, we get better tools and a better 'overall' engine.

Ron

#6 Ron, for the record, it doesn't help you, and we never claimed it would :P. 64-bit lets the executable use more RAM.

Quote:Perhaps what must be done is create new functions that is general and will cater to all game play genres, and expose some of that with torquescript.

Yep, that's exactly the plan. Also, the engine needs to come with a non-FPS template, so people can start building off that instead of trying to disentangle fps-oriented stuff from the current templates.

#7 @Ron I'm still a new settler and "testing the waters" on this community so I have no idea what has happened on the past. it's a shame you have to leave the committee but i'm sure for some good reason. I think dev teams for open source project will benefit if you still are or have another artist on board.


for me as an artist just starting to get a grasp of the engine, here are some major technical issues I am currently dealing with in creating a decent scene

Lighting:
global Illumination (Static or Dynamic) - whatever mode it is, GI is not a luxury feature anymore. It is a standard nowadays to make visually appealing games or just scenes. I don't see any dynamic GI plans at the moment or internal lightmapper cause I know this will require heavy shader knowledge from a contributor but still we can make use of static option with 3rd party tools like blender/ but the pipeline connecting to the engine is broken. I have indicated this on my first post "Lightmaps and Dynamic Shadows"

Anti-aliasing/Final Render:
viewing my baked scene to another tool like blender or other game engine looks ok, anti-aliased but still crisp good lighting. but when I put it on torque. it looks blurry with only 1x anti-aliasing or off. what's the deal here? must be the shaders or just properly setting it up, i dont know..

Documented features with no live examples:
this has been discussed. just pointing out here.

I wish I can do something about some of these items or to at least contribute but I'm not a shader expert or coder. im just the average artist who knows a bit of scripting and modding and maybe later on can put together some nice demo or hopefully a game using torque.

#8 yes totally agree with Ron, for the last year that I'm here, I can remember and celebrate acceleration only two aspects of the set that even as it changed the engine, is the addition of tape instead of particles, and then the water stopped making waves with superfluous by turning the camera on it and it's just a drop in the ocean of the changes that have been made but the rest is almost no way affect the common opportunities, all dance around a lot of shaders and other "dermis" To get candy for the eyes, but it is not worth a brass farthing, without real (even if many bugs) but innovations and optimization work, the user as a builder and effectiveness of the computer itself (torque 3D is the only program of all that is on my computer overheats it to such an extent that it crashes after 15-20 chopped minutes of work with this enabled) I have more than a year dancing with a tambourine around documentation and other tutorials that with nails nailed to the shooter tutorial, and I still could not give a character in the hands or where any node to at least one thing, it crumbles my head, no matter what (even a lot of effort on the steering committee) torque still look even just unfriendly, it looks dangerous! and aggressive for me :(
(sorry for these cruel words, but it's true, by the fact that despite all the aspirations of the community and to bring something new torque does not change seriously, but just think what he pretends to be makes an effort), I'm still waiting and hoping for improvement is not losing faith, I now powerless to help to you sleeves rolled up and diving into the code head because my qualifications just simply is not enough at the moment, but I'll just conjure, add and add and add again! Merge into one canvas all the colors, all that can only add, truth is (if the jar is empty and you have water in the sand. how would you not shaking it, clean water in the jar will not appear on this Fill the jar dirty water with sand, wait until the sand settles to the bottom of jars, and the top will get water without sand) do not be afraid to dirty jar with sand, afraid to die of thirst. that I am say anything Depends on my, I call a steering committee to arrange Hackathon, and in the shortest possible time to add all the tools that gave the community (such as ... volumetric fog, the door to find a way, an updated interface, skin, etc, etc. simply collect all of them in a pile), that is, those that people from the community were released under the MIT or a compatible license, to make it simply is working and give a quick shot, and only later to rub to a shine with all the time.

#9 I think the developers are moving the correct way, fix bugs and then make new features. I also think the issue is not the template but the lack of documentation on how to argue with something that's not the template: btw also documentation is growing right now.
Personally, I think everyone embracing T3D should share his improvements and fixes, that could break things but involves the user to have the ability to code and adapt and learn to patch on need.
I like Torque3D because I don't have to look for plugins, templates or extensions: if you need something you can create it. It's time consuming but is a one-time job if community is active and you share.
For example, I'm making a dogfight game, I did a basic radar, fixed theora bug, fixed waterplane glitch, added aircraft roll etc..those fixes are working for me but themmost are game dependent, but if someone comes to ask if I will share them the answer is absolutely yes. The reason why I'm not doing a pull request for my stuff is not every game is equal to the others and IMHO best thing to do with T3D right now is doing what they are already doing: cleaning up, fixing and then add features. I don't want to end up some day with some kind of "engine" where everything is drag and drop and if you have something particular to do you have to pay for a pack. I'm happy to have the precious opportunity to dig in code and do what I want. Is not user friendly but is absolutely freedom friendly.

To make it short, the guys on github IMHO are making an awesome job, no need to lose time making templates (now) and the only thing T3D really is in need is more documentation and love.

#10 Okay, if we're looking to make a FPS template, lets dust off the FPS Tutorial and make it a template. The documentation needs to be updated and some script fixes are probably required, but a large portion of the "how do I do this stuff" is already there.

I like the idea of creating templates, but I'm not sure I like the thought of spending resources there when there are bigger fish to catch. Piggy-backing off of the FPS Tutorial saves a ton of work.

#11 Well, this thread went all over the place.

So, the point is "we need better, fleshed out examples that are more like mini games" of different genres and gameplay types, rather than half-baked templates?

If so, I'd agree to that. Richard's point is very salient to this.

Having something that's a miniature game on it's own would be a better starting point for people developing as well as draw attention, and having different types of these is important for us to know what areas need improvement for non-FPS types.

I've brainstormed some in this aspect, so if this is what people are more generally wanting, I'll put a bit more polish to the idea and we can build a plan around it.

#12 Kind of what I planned to do for 7DFPS, but without the baggage of working off the current templates. Also had plans to start on a Zelda-ish template at some point- Jeff and I have discussed this.

GamerX and Olexiy, while I appreciate the problems you guys are having, and I definitely share your pain, we seem to be diverging back onto the 'artists and level designers' thread, so let's keep things focused here.

Am I sensing that we might finally have enough momentum to start a 'community game' sort of thing? I've always believed that the community isn't currently large or organised enough to pull that off, but the idea keeps coming up.

#13 thanx for the feedback guys. ok understood. at least start first with the live basic examples then the move forward with full blown templates or demos you are planning.

as long as I or other beginners get to try all torque features with ease, we're good to go.

#14 Instead of trying to make T3D a general game engine, which it is anyway for any programmer with the ability, concentrate on its strengths. There are new game engines that are pushing themselves as FPS makers, similar to how RPGMaker is so clearly branded.

Maybe just rebrand T3D in a similar way, maybe Torque3D the FPSmachine or something. I put this forward because I have all the Torque engines from before it went MIT and I now use other stuff as well so I see a lot of what comes out, as surely you all do as well, but never mention. The houdini engine has been released for both UE4 and Unity, Amplify texture 2, terrain composer and the list goes on. UE4 are now positioning themselves as a general game engine with blueprint, so to me it seems like an opportunity for T3D to recapture the FPS space. No-one is saying don't put templates out there for other games as well, but if FPS is the main focus at first then one is building on a strong foundation and one could add stuff to the engine focussed on helping FPS developers do a more polished game.

A more narrow focus rather than a broader one might be better, T3D cannot compete with either UE4 or Unity as a general game engine because both of those engines have 100+ developers working full-time on them, not to mention Torques codebase is already built around FPS.

Instead of trying to make Torque what it isn't make it better at what it is.

#15 I for one think torque should take on a more plugin based approach, drop in drop out features are a big hit but to create that would strip the entire engine back to it's platform layer all other features would be optional. But then again we all have the source code - don't like something then change it

#16 For the templates. I feel multiple different templates/genre kits would be the best way to go. That's what the original torque had going and one of the reason me and my college buddies was attracted to it. I think that concept should be brought back. Like Ron and others said maybe a few different small groups could get together and flesh out these different genre type demo complete with scripts, art, and etc...To give the user a good starting point on the specific project they are working on..
I've been saying this for years and I also mention the use of maybe using garage games models from the past but I recall them saying there would be legal issues. So I would suggest lets just recreate some of those. I would most definitely be interested in join a team or two, to help get these going. Looking at Ron ;)
Anyways with the other project(s) I was working on with my brother nearing completion and my project currently on hold for a short period I have time on my hand so anyone interested let get together and come up with something.
Now on a side topic "tools". Personally I believe the engine is complete other than having a solid out of the box AI solution and a cinematic tool (IMO). Now there is Daniel's walkabout that is now open source (again Daniel thanks for this)that would help on the navigation part of the AI bots. Now we would just need something that makes setting up different AI types or brain types easy, something similar to uaisk and also including an editor would make things easy for someone that is just pick up T3D and get going. The cinematic tool I would easily throw my suggestion of Verve for that. I haven't gotten around to trying it would my self be from what I seen it would be a great addition to the core engine (imo it should have already been added).
Question to Ron, have you had a chance to try Verve out in your personal build, and how does it do? Is it easy or take some time to learn?
These two tool sets would help make things easy for packaging game specific script for these genre kits. People would be able to package verve specific camera cutscenes, and paths scripts, and etc from verve and whatever AI solution's default AI packaged along with a "howto achieve docs/tutorials" of the various different things in the kits. That's just my two cents!!

#17 I have a few things that I have developed over the last year ready to deploy few things I promised almost a year ago has took me this long to get it all done lol implemented a few cool things that should help everyone out like have been really quiet to get them done but coming months should release everything, anyone needs shaders, modular weapons, better postfx better DOF (with all be it a rather poor bokeh effect) and hit boxes, also aiming nodes for Ai so you can say right Ai aim at this node instead of his feet, also few other things for an RTS type game

#18 @GamerX

Quote:Lighting:
global Illumination (Static or Dynamic) - whatever mode it is, GI is not a luxury feature anymore. It is a standard nowadays to make visually appealing games or just scenes. I don't see any dynamic GI plans at the moment or internal lightmapper cause I know this will require heavy shader knowledge from a contributor but still we can make use of static option with 3rd party tools like blender/ but the pipeline connecting to the engine is broken. I have indicated this on my first post "Lightmaps and Dynamic Shadows"

I agree I do believe there are a few branches attempting to update the lighting system. The deferred shading for one. They are doing that in an attempt change to "Physical Based lighting" :)
Also currently Lukas created SSGI lighting that can go into the current version of T3D (which I'll be adding to my personal build). I believe he was attempting to create the same effect Ron was a few years back with single bounce global Illumination. I think that would be a good addition. I haven't gotten a chance to try it yet myself but if everything turns out good would make for a great addition. I think his currently overrides the SSAO but if we could find away to have both it would give the user a variety to choose from.

*EDITED* I also think the technique Felix (FEWES) did for the eye adaptions and HDR would be a good addition to the engine. I now understand it wasn't a new shader he did but just a hack to the way indoor and outdoor scene are read.

Quote:I extended the zone object to include an "Indoors" value which could be set on any given zone. This value was then registered as a post shader variable and then used in the luminance pass of the stock HDR shader to brighten the lighting depending on if the camera was indoors or not.

@Thomas, I personally would be interested thanks!!

#19 yet another pointless debate with pointless opinions made mostly by people who dont/cant/wont do anything to contribute, I will also add myself to this pile under the dont and cant categories.

Let me start by invalidating the opinions of most of the people (in my opinion, take that anyhow you like, I dont care enough to be offended).

Ron, not to actually pick on you, but you cannot compare blender project to the torque project, blender has goodness knows how many dedicated and paid developers as well as dedicated and paid other company positions, the software maybe free and open source but for the most part its being developed by a corporation.

Anybody who says 'you have the source do what you like' is somebody whos opinion is automatically invalid in discussions like this, not everyone has the ability to edit the source code in all the ways needed to change the course of the engine.

Anybody who says engine xyz has this feature, feel free to take the advanced source code out of any other free and open source engine thats beter than torque and implement it into torque, do not compare torque to multi million $ engines (Unreal, cryengine), do not even compare torque to unity since thay have enough purchasing power to "freely" hand their users $,000 upon $,000 worth of middle-ware which in real terms probably would cost $x00,000's f you wanted a source solution to add to an engine. Your opinions are also mostly invalid, if this is your only reason for wanting xyz feature all you are saying is "i want a $million engine, for free with source code"

The main problem with open forums is that everyone wants different things, the real problem is that this is an impossible task given the nature of the beast. We have people who don't even want to see source code much less touch it. We have people who don't even want to see art, much less touch it. and in two sentences the scale of the issue becomes quickly apparent.

For a long time now, i have been a vocal proponent of a number of things.
1) that we need 'shiney' demos to attract people here
2) we need usability to keep them here
3) we need sample genre templates for at least 3 major genres, FPS, RTS, RPG

The dedicated (as in i loove torque) programmers in IRC from both the SC and the non SC pools are adding all thefeatures we need to make item 1 a reality, what they need is some backup in the form of artists and level designers to build and show off some scenes, even if these scene are only possible to run in quad titan type computers (much like unreal promo.demo vids)

The usability angle requires artists that do know how to code to explain why the editor isnt working for them instead of just saying how hard and awkward it is (eg non x, y, z rotation values in the editors being majorly non intuitive).
Heck, some of the artists for sure know enough coding to fix the UI themselves but sit on the belligerence bus.

and item 3, regardless how they are added or presented, how they are incorporated into the base engine, creating these templates with levels that show of how feature a b c are added and working will invariably highlight any parts of the core engine that dont easily work in favour of that particular genre, and example here is camera controls, most can be easily done, but examples are few and far between, some may been some source tweaks to make the cameras all things to all games.

As far as development is concerned, we are currently stuck with a steering committee model that works in theory and doesnt always work in practice.

The engine for the most part only advances when somebody who is using the engine to make their own game adds feature d e and f to the engine, nobody is actually working on the engine for the sake of working on the engine, which means the rest of the users of the engine are stuck with whatever features are being added to other games. Dont get me wrong some of these are awesome additions to the engine, but I think that is more of a good fortune event (for the engine)

Getting around to the misconceptions of the original post, the FPS template isnt even an FPS template, and the artists tools are not anywhere near as bad as they are painted if you will excuse the pun. Sure it could be improved but its not like its unworkable. As for changing the engine to yet another FPS engine amongst all the other FPS engines, that frankly is just self centred thinking.

#20 Bloodknight
I agree with you for the most part. I've been suggesting the same things you've mentioned above way before their was a SC, when T3D had a company behind them and still developing T3D. I also suggested putting my different project(s) aside to help contribute to creating the genre ports/templates but I never get any replies (still willing by the way)..I would do them myself if my skill base was there but unfortunately it's not!
If I was forced to do it my self then I would be force to charge for the work I put in and I don't want to do that especially considering all the knowledgeable people here and could contribute. Some people even have things that would be beneficial to the engine but for some reason refuse to release it or just sit on the tech..I know me for one I have screenspace lightning and a realistic skin shader that I would love to contribute but I wasn't the person whom made them and under his license I can't..I'll continue to ask him though.
Nonetheless these thing aren't going to get done unless people get together and make it happen..Like I said I'm willing to help if I can get people willing to contribute if not then I fear this engine will always be in the position that it's in.
People suggesting good ideas but NO one willing to make things happen!!!