Building our own 3D engine

[Frank]

Do we want to? And if so, who would like to participate?

 

[digitalwanderer]

Why not do as you suggested and just use Ogre? Modifying it for our game should be enough of a project as is, at least it'd be a good skeleton to start on.

 

[Frank]

Why not do as you suggested and just use Ogre? Modifying it for our game should be enough of a project as is, at least it'd be a good skeleton to start on.

Yes, I agree. But I want to cover all the ground before we really start with it. I build several (2D) game engines in the past, but we need more than only that. At least 10 people, and the art and level design is the hardest.

The engine by itself is something two people can do, while all the others build the content. And a brand new one can be the major focus point on a board like this.

 

[Rys]

I've seen the other thread, and while I largely don't have the time right now to seriously help, I am willing to provide resources at B3D to make it happen. Server space, code repos, private forums, etc. Let me know what you need, DiGuru.

Looking forward to helping out on the coding side when things calm down a little

 

[Frank]

I've seen the other thread, and while I largely don't have the time right now to seriously help, I am willing to provide resources at B3D to make it happen. Server space, code repos, private forums, etc. Let me know what you need, DiGuru.

Looking forward to helping out on the coding side when things calm down a little

Thanks! Gladly accepted.

Unless people prefer otherwise, I am suggesting we use Ogre and start building our virtual playground, residing on the B3D server.

An announcement and dedicated forum would be very sweet. That makes it official, and will help very much.

Very nice!


Edit 2: A source repository from the start would be great as well, do you have one in mind? Otherwise I'll look into it.



Btw, that's what you get from answering while doing other things: you miss half of it.

 

[digitalwanderer]

Yayzers Rys! Thanks, damned glad to hear you're on board with supporting this.

I'm looking at this as the ultimate game design learning experience, I'm very excited.

 

[ERP]

FWIW I'm building something at an incredibly slow rate on my Web page, more as an architectural excercise than anything really usefull. And I only update it when I'm bored or have a couple of days off.

If you really want to build a game engine, build a game and use ti for your requirements.

My experience is that "Indy" developers tend to get sidetracked by technology rather than working on the game. That's fine to a point, but if you want to build a game, build a game and beg, steal, borrow or just hack the technology you need to get it running.

 

[Razor1]

hey if ya guys really want to make a game, I've got a game for ya, we are short of coders, using the cry engine (fully licenseed), I'm the only level designer and level asset maker, but I have it down to a T takes me about 2 months to make all level assets per level (most of the time takes when rendering out high res textures). Would like help with texturing though, I'm good with using vertex paint to get the base textures done, very high detail, but need final touches. Characters for our demo are all done, being animated now, also we have a few guns and weapons, but once I"m done with the level, I will be making more.

Edit:

Game design is pretty much complete. Still some loose ends on the story and melee combat system though.

Also I'm good with making art assets, very good, but this will be the first level I've ever built, I understand the cry engine very well, so lighting is probably my forte with level designing, but level layout, I'm not very good at, getting better, but its a bit hard without much background. Engine coding isn't really neccessary for the project as most of the modifications are completed, by me and our lead programmer, but still a few minor modifications still left, mainly with physics and ai.

We do have a somewhat strict schedule, but nothing concrete, of course faster the better anyways, if you guys are interested post here, or pm.

 

[darkblu]

i think people need to decide whether they'll be building an engine or a game. those two things can take place concurrently, but the focus should be clear - getting a playable game done, or getting an engine with a semi-playable tech demo done. and this focus should be clear to every participant on the project.

oh, btw, regarding ogre - it's a rather fat piece of code, if you're in for the 'my engine' experience, starting off with ogre may not be necesserily the best approach to get a grasp on concepts. i.e. at the end you'll eventually know how ogre works, not necesserily how to create a good engine meeting certain needs.

if you really want to start off with something in place, and not a clean sheet design, then i suggest you take advantage of some experienced game/engine dev's hobby proj, like ERP's.

 

[K.I.L.E.R]

I'm more than happy to help out on small things provided you give me the documentation which states the requirements and such.
Sorry, it's the best I can do at the moment due to work.

 

[[maven]]

I may be able to contribute some code / tips, but I'll have to check with management...
Though not too many people are doing software rendering / clipping these days... (other than the Assembly'06 winning demo) .

 

[Rodéric]

Though not too many people are doing software rendering / clipping these days... (other than the Assembly'06 winning demo) .

Could be nice to have a very fast software renderer for occlusion, would avoid many troubles of trying to use the hardware for that, and make the engine generic enough.

 

[pascal]

My only advice if you develop a new engine is use a multiplataform open API like OpenGL, and have a port to Linux in mind

 

[Arun]

I wouldn't be against helping develop something in-house, if we could GPL it and have high quality standards. Developing an engine ourselves is only good if it's better than what we can find elsewhere for free. And if it's better, it doesn't make sense to keep it for ourselves.

Ingenu's engine is another option, which has the advantage of being developped by a B3D member, even if it's maybe not exactly what you have in mind when it comes to developping our own engine.


Uttar

 

[Rys]

Developing an engine ourselves is only good if it's better than what we can find elsewhere for free.

I'm not sure I buy that. If the whole exercise is one for learning, then surely the experience itself is worth something, even if the finished engine or game isn't brilliant, or even any good!

The engine doesn't have to be UE3-beating, as long as the developers of the game are happy with it and get something significant out of it and the experience in building and using it.

IMO.

 

[Rodéric]

Yep really depends about the goal, if you want to create a game using an existing engine is much better, you can tweak/change it all you like, but at least you'll have something that works (my engine, *hint*, *hint*); if the goal is to write a 3D engine, then you'd want to start from scratch and design it from the ground up, I'm sure several people around here can help with the design, even though I bet we would disagree on some design choices ^^

So is the goal to get a game done, or to learn how to write one from scratch ?

 

[Razor1]

Very true Rys, the best way to understand and learn about 3d engines is to make one! Its a very gratifing feeling when the first game code is put into action too

Oh yes and its definitly possible to build an engine that is comparitive to UE3 or any marketable engine, just takes time and very good documentation, and a clear plan. (there are alot of free software packages that can be integrated into a 3d engine to make it close to full functionality of the UE3 engine, I would say other then the graphics core and AI (there are some rudementary AI engines out there but they need quite a bit of work) everything else can be added on from GNU/GPL projects)

 

[Arun]

I'm not sure I buy that. If the whole exercise is one for learning, then surely the experience itself is worth something, even if the finished engine or game isn't brilliant, or even any good!

Fair enough, I should have made that sentence clearer. IF the point is to make a game that hopefully is any good, then developping our own engine is going to make that harder (by diverting our resources) if it isn't actually better than what you can get for free otherwise. If we see it more as an intellectual exercise, that's completely different of course.

Uttar

 

[K.I.L.E.R]

I believe that a pre-existing product can be effectively put to good use and extended if the learning curve isn't as substantial as creating a new engine.

Another thing to factor in is how abstract will the engine be?
In the end you will either want the engine to be suitable for a certain game or do you want to create something that can be used for anything anywhere. The later is certainly harder as you would need a lot of time and manpower but most of all you will need to spend a large amount of time planning how you will implement your features. You would want your features to perform as best as possible under different configurations and a problem that comes from this will be the amount of testing needed to sort out any performance issues.

 

[NRP]

The reason I originally came to B3D and the reason I stick around is the same: To Learn. This place is loaded with knowledgeable people, and I think this exercise is a wonderful opportunity for learning.

I understand that some here are industry professionals who have been down this path many times before, and are not so enthusiastic about going down it again. Nevertheless, I vote for the main goal of this project to be learning.