What would be the best game development tools, in your opinion?

[Stev Studio]

Hi,
My name is Steve. I was working for an Activision studio as an AI programmer here in Quebec, Canada. I recently quit my job, to work for a smaller company & start my own.

I am currently working on game development tools that I plan to release commercially in 5-6 months.

I am working on a game engine, a level editor, and plugins for various softwares, including 3ds max, maya & adobe illustrator.

I want to create a very easy to use asset manipulation process, so developers can work on their games instead of their technology & tools.

I would like to know, which features do you think would make my game engine the best, in your opinion?

Thanks for sharing your opinions,
Stev Studio

[jankoM]

various softwares, including 3ds max, maya & adobe illustrator.

One feature I don't like and many of these smaller/new 3d engines have is that a lot of them suport 3ds, maya from the start.... and then in their fifth plan some will support more accessible tools like blender, wings3d...

//edit : why would I but $200 if I can afford tools that cost several $1000? I made similar comment at one nice 3d engine which said that you need just notepad to start making a game... well they forgot they can only import maya and 3ds so you need something a little bigger than notepad too

[PoV]

I recently quit my job, to work for a smaller company & start my own.
Sounds familiar. Welcome to the party.

Making a business of tools and middleware isn't a bad idea, but like any successful product, it'll have to stand out to be one. So my advice, specialize. Know you're competition, see what they don't do so well, and if there's enough of a demand to make it viable, do it. Ask who could use your product, and who will. Middleware, with the exception of a few game building suites, isn't all that popular outside retail development. The tag line "buy my engine or editor" doesn't give me much to latch on to as a consumer. Assets/Workflow is a good start, but you have to understand that most people don't understand why workflow is important. So if you pull it off, you'll have to find the right words, publicity and examples to convince them.

And like jankoM pointed out, support for budget tools like Blender are important in this space. Even Softimage XSI, despite Max and Maya being the defacto at most studios, XSI is only $500 for a usable license.

Good luck.

[esrix]

If you want to reach a broader developers market, making your engine cross platform would definitely help. Many developers use Linux and some even use Macs. Take open source APIs like Irrlicht, ClanLib, SDL, and Allegro for example.

[mot]

Crossplatform (Windows and Mac), allow custom output formats for the editors, very, very good examples (HGE had good ones), everything simple & solid (cool shader effects are nice but not crashing is more important).

When I look at irrlicht, the only thing I can imagine using it for is a FPS. Ogre3d is HUGE. Open scene graph is also interesting but too hard to set up and maintain. A simple precompiled library with little or no external dependencies would be nice. Keep all the tools and documentation up to date with the engine.

Make it easy with the 3dsmax and maya plugins to include extra "meta" information into the models/maps. Example: let's say I want to make a racing game, something like Need for Speed. So I need to build the tracks and place special items, hints for AI drivers, start/finish locations and other information into them; not just as the last step but repeatedly, I keep tweaking the map and the items have to change as well. Makes sense?

Also I will want to test the maps quickly, so the time between saving it and being able to view/play it in the game should be very short.

[esrix]

A simple precompiled library with little or no external dependencies would be nice. Keep all the tools and documentation up to date with the engine.


That reminds me; The Apolcalx Engine (http://apocalyx.sourceforge.net/) is a precompiled engine that, when started, looks for Lua scripts and executes the chosen script. Entire games are often written in the Lua scripts and executed by the stand alone engine. Unfortunately, the engine has a habit of crashing on certain games depending on the user's hardware.

[Sharpfish]

Everything above... but to restress: Easy X-Platform (Mac/PC).

[zoombapup]

My advice, not wanting to sound negative.

DONT DO IT!

Every programmer in existance starts out writing "their engine", then they progress from there to nothing.

Believe me when I say that this happens ALL the time. Its the NIH syndrome writ large.

If you prefer making tools/tech rather than games. Then make your middleware targetted and easy to understand. Good examples might be speedtree and pathengine. Both are relatively "black box" from the customers standpoint. Want trees? speedtree. Want pathfinding? pathengine. See?

So find something that isnt easily solved already, and solve it. Just one problem thats nice and useful.

For instance, 2D physics engine. Or a flight model for a plane (hehehe.. just random thoughts honest).

I'd much rather license a robust single purpose toolset for a given problem, than yet another "engine" which tries to do everything and ends up doing nothing really well.

If you DO go down the engine route, my advice is to basically do a 2D engine like PTK. Look at how popular that is on these boards, given there is virtually nothing to it (from a high tech standpoint, I'm aware it may have taken a lot fo work).

How about this for an idea. A 2d "scene editor" that works a bit like worldcraft but in 2D. What it does, is allow you to place sprites and tiles, but based on a timeline event based system. Such that say you were doing a shooting game, you'd have a level editor that allowed you to spawn the various enemies (not hard coded, but using a "fgd" type thing), but also allow you to spawn them *at time x* and the ability to place movement splines and etc etc.

Basically an openly scriptable/extensible 2d scene editor.

Ok, thats my 2 cents.

[jankoM]

Such that say you were doing a shooting game, you'd have a level editor that allowed you to spawn the various enemies (not hard coded, but using a "fgd" type thing), but also allow you to spawn them *at time x* and the ability to place movement splines and etc etc.

Hmm.. this is allready done by the same person as PTK :) http://www.phelios.com/splined.gif , http://www.shmup-dev.com/forum/index.php?topic=232.0

[svero]

Id like to echo zoombapup's comments. I think he's got it right. This sounds like a lot of effort for very little reward. If you are committed to making tools then I'd suggest a niche market, but even then I can't think off too many game tools I'd want. Mostly I just use the standard apps that are out there like softimage, 3d studio, paint shop, photoshop, etc... In terms of engines there are a bunch already and some of them are totally free. About the only thing I can think of that I'd buy immediately is something that makes it trivial for me to make flash games by coding in C++. I can't stand actionscript, but maybe I'm not the norm on that. I've seen a few c based libs that are supposed to spit out swf, but nothing complete that works well as of yet.

[MrPhil]

If you DO go down the engine route, my advice is to basically do a 2D engine like PTK. Look at how popular that is on these boards, given there is virtually nothing to it (from a high tech standpoint, I'm aware it may have taken a lot fo work)

I've heard more than one request for a GUI system for PTK, just a thought.

[luggage]

A good quality GUI system would be useful. I had one I used a while ago which would work with practically any engine but it's suffered code rot.

[zoombapup]

Janko: I meant a bit more polished and "open" than just a windows apps that places splines. I'm thinking of something with a timeline etc etc..