Ogre or Irrlicht

Well I suppose it is time to start this discussion. For the MMF 2 3d object, do you guys think that Ogre or Irrlicht should be used?

Both are 3d engines and both are open source.

Ogre 3d
Irrlicht

Irrlicht is really easy to use, but Ogre has more flexibility. I think the ease of use would make an MMF 2 integration somewhat easier but maybe not.

I want to get a feel for which you guys would like to see. Especially if you have experience in either of them.

Before you comment on one or the other please take the time to check them both out and see which you like the best.

Also is there interest in helping me out with this project? My freetime is painfully short and while I would like to kickstart this project, it might be up to you all to finish it and get it out the door.

Thanks.

The general "buzz" I've heard around the internet sais that Ogre is better. But I have no personal experience to contribute with ;P

After viewing the screenshots of both, I must say that Ogre looks amazingly cool and realistic.
I believe Orge would be a nicer engine to use.
The ease of use depends on the way you make it. :)

I have to agree with LIJI. From a superficial standpoint, the projects done with Ogre look very professional, while all the Irrlicht projects have this "homegrown" look to them.
Of course that heavily depends on the people using the engines - but then why do the pro's all chose Ogre?

What happened to TurboFerret's Irrlicht extension? Didn't he like.. do half the job already (if not more)?

I have worked with Irrlicht a little and I've found that it's a little too limited as it (no MD3 support for example)... Ogre seems to be difficult to use but it seems to have better functionnalities.

Ogre does seem to be the better engine, but it would be cool to have the power of either accessible in MMF2. But are you sure you can take all this stuff on? You've got loads of great stuff on the go, I'd love to see you get the physics extension completed. I wouldn't want you to burn out taking too much on :)

This extension is going to be a team project, I hope:). At least it was discussed before... So lil' Vortex wont explode:)

Edit: My vote goes for Orge

Just from the screenshots I would say that Ogre does look like it can do great results.

I agree. After looking it over, Ogre is looking good. There is some nice stuff that people have made.

Seems like Ogre is a very popular response :).

As far as the Irrlicht object that Turboferret was working on, I have no idea what happened to it or where he went. He sort of just fell off of the face of the earth.

My plan is to have a very different type of mentality in terms of the 3d extension and that is that I am going to make the entire project OPEN SOURCE to any developer who has shown merit here in the EDT and wants to help work on it. I am going to look into the logistics of setting up some source of private sourceforge type thing where I can control who is granted access and people can upload their work.

The way I see it is that this project is too much work for one person so making it a group project would spread the work out, make sure that we don't have to start over again, and it will ensure that the project will hopefully continue to develope even i I don't have the time to constantly work on it ;) .

I personally would like to see an EDT attempt at 3D which perhaps could be partly sponsered by CT, I am sure CT UK could put something towards it, its about time i sponsered something for once :).

I wouldn't personally like the idea of it being restricted to only developers you allow access. I think anyone in EDT has already shown there merit.

I think it would be good if CT were heading it up as well as this is a biggie, perhaps Rhon would like to pick it up and project manage it?

I just think if something this big is going to be taken on, it would be good that there was some reward, even if it was only little.

Anyway let me know your thoughts, and any other ED's, on how it could work.

Jason

I think I should be allowed to manage my own project. I think it is wise for a developer to lead a development project...

Of course I would let developers from here into the project, I would just like to screen them as a form of quality control to ensure that they have some experience working with Ogre first... That means that I would just like to see a small application that uses Ogre. This is the same thing that I do myself before I create any extension is I make a small program using the API to get familar with it.

I also want to design the extension in detail before a line of code is written.

Im not trying to take take over your project or anything like that, theres nothing to stop you from working on Ogre, but i think something like this requires a group effort, and I personally like the idea of something like this being a CT led project, because it means that it can be more focussed in a particular direction and there is a possibility of payment for people involved. I think this way there is more chance that it wont disappear into the abys, like the Irrlicht object because people have other things in life to do :(

You may not agree, and thats perfectly fine. But im just putting it out there for discussion for you and the other ED's.

Thanks

Jason

I don't disagree with the fact that it should be CT sponsored and have CT involvement but I also think that I should be allowed to direct the developments to ensure that the project develops.

From your post it certainly sounded like CT wanted to fully take over the development anyway. I personally do not agree with this and furthermore I am not going to allow it.

Thank you.

Let me clarify. It would be easier for someone from CT to push the extension into a direction that would not be the best direction for it to be pushed into. Unless the leader of the project had sound 3d experience then to have them make decisions about what features to have for instance is a bit rediculous.

I am worried that if it were CT lead by anyone other then Yves or Francois then it could end up suffering from too many features that are not really needed and too few features that are.

Of course CT will get a lot of input into the development process but I want to be able to make the calls as I DO have 3d experience.

Quote:

---

As far as the Irrlicht object that Turboferret was working on, I have no idea what happened to it or where he went. He sort of just fell off of the face of the earth.

---

I think he got discouraged because there were so few people with both the 3D experience and the time available to test out his stuff. Maybe Andos has news of him?

I know 3D would be nice, but I do wonder if we have sufficient demand and experience in the community to take real advantage of it. You'd have to be very patient and be prepared for it to be a slow burner.

AndyH has masses of relevant design and development experience, but he's got lots of other demands on his time.

I do wonder that myself. It is a good point.

I think that with the proper tutorials and resources, plus the fact that it is 3d, will inspire developers to learn it.

People used the 3d mesh object, Mode 7, KlikOpenGL, ect. They even made some neat things in it despite the fact that they were a bit weak. To me that says that there are people in the community who would be able to use this stuff and to take full advantage of it :).

I would say, im not here to steal your work or take over your development ideas.

I fully understand you want to do this extension, and its open source, so no one can stop this, or in fact prevent anyone else giving it ago.

I know how busy you are, obviously your still working on the Physics object which has taken awhile, and I worry this extension will take a very long time without someone pushing it, but without it being a group effort its not going to happen.

Im just trying to add my knowledge to the post, and looking for ideas and comments, im not trying to upset anyone or prevent you from doing what you want. But I can see there is a big possibility that this would fail early on if the lessons are not learnt from Irlicht.

Thanks

Jason

I'm agree with you Vortex2 and to be sure about that, I've made a little poll on a French website about development priority of MMF2 and 3Dobject is the second one after Hardware acceleration and before Vitalize4...

So I think if a 3D object is developed for MMF2, a lot of users will try it and some developers will work for it (if they can)!

EDIT: But I think Physics has a highest priority for the moment. :)

Well, I made some pretty nice stuff in Jamagic too, but you don't get showcase games done without a dedicated team. The graphics are the killer - you have to have lots of prebuilt models for people to use.

I'm not saying it won't happen, but be prepared for a lot of frustration at the slowness of the uptake. Are you sure your time wouldn't be better spent on some of your other projects? You were getting masses of forum feedback and anticipation for the Physics engine.

I was actually talking about XNA:

screenshot

but anyway I am not out to have an argument. Anyone is welcome to give a 3d object a go but this remains my project and it isn't open to discussion. If CT wants to lead a different 3d project then that is up to them, and if anyone wants to help me on mine then they are welcome to.

Thank you and I hope everyone understands my position on this :) .

I'm missing some vital information here [smile] What is XNA when it's at home?

Its a game development product from Microsoft

Indeed. It is an API from microsoft in the C# language that is built off of DirectX for the creation of windows and XBox 360 games.

Basically in practice it is just a layer above DirectX and it doesn't have the luxury of a premade engine like Ogre (yet). I had to create the 3d engine from scratch.

Anyway we still have quite a bit of work to do before development begins anyway. For starters I have to get all of my current extensions out the door and that will happen very soon.

Then we need to start getting ideas together and formulating a solid design. The extension will be designed before any c++ work is done.

I'm still feeling incredibly dense this morning. [blush] What has XNA do with the graphics provision issue, or with what you are proposing for 3D? Or maybe you were answering someone else's question instead [laugh]

Tiles has proposed him to help about 3D lowpoly graphics :)

I'll can probably help for the extension or for somes plugins of your 3D extension Vortex2. ;)

It was meant as an example of 3d experience outside of Jamagic and because I wanted to show off the screenshot XD. Nothing more really.

I think the logical place to start on all of this would be to compile a list of all of the 3d suggestions that people have had and to try and figure out a good way to fufill most of them.

Any implementation we do needs to be powerful enough to allow for a wide range of games and yet not be overly hard to use. Some simplifications will need to be made I am sure but what they are remains the question.

That sounds awesome, thanks Sphax and Tiles :) .

I'll can probably help for 3D lowpoly graphics too. :D

Finish your Physics object before and after you can go on your 3D extension. lol. ;)

oh of course :). Because after the extension is created I will need to create a 3d physics extension :D .

Sounds too exciting :) If I could post my wishlist for your time and effort, it would be 2D Physics, 3D extension, 3D phsyics extension :D

Just started to play with ogre, and I must say it is very interesting:) specially the particle scripts.

Awesome :).

There are a number of things that need to be considered when working with 3d and when designing this object.

The first thing that we need is the "core". The core object is the object that is responsible for the initalization of the Ogre engine, allows us to choose a rendering subsystem, holds our "Scene Manager", ect.

The Scene Manager is what manages our cameras, lights, objects, ect. It is the main thing that is needed to create an Ogre scene. The "core" would allow us to interact ith a scene manager.

The next thing to consider is the "Entity" class verus the the "SceneNode". The Entity class contains 3d meshes and the "SceneNode" class does the location stuff. This is going to be a tricky thing to implement and might involve some innovation to get correct. I think we need to combine Entity and SceneNode together even though Ogre keeps them seperate. In this case the main component would be a SceneNode and we need to deal with allowing for multiple Entitys per SceneNode.

The Camera is the next thing. We need to only have one camera but to move around to different parts (to make it seem like we have multiple cameras) we would need to be able to move to various scene nodes. This is how Ogre is somewhat different then other engines.

The camera class would need the ability to look at objects, ect.

Viewports are a bit tricky but they could allow for multiple views for like split screen and stuff. Not sure if we are going to implement this feature or not.

Then we come to lights. These are Point, Spotlight, and Directional. We should provide this as one Light object and allow for the user to select which type they want.

Shadows are handled by the scene manager and we need the option to be able to turn them on and off and select what type of shadow we want such as Modulative Texture Shadows, Modulative Stencil Shadows , and Additive Stencil Shadows

Fancy elements like Terrain, Sky boxes, and Fog are all nice but they will not be implement right away. We will focus on the core components first and getting them working.

I want to try and provide the full range of GUI components and the ability to load layouts.

Then there are a ton of math related things, animation (bones), ect. It is a lot to cover!

The first stage in production I would like to see the "core", the "camera" and the "object" extensions created.

The core is the most important as it ties everything together and constructs the scene, the camera lets us view the scene, and the objects allow us to place objects in the scene.

Using these three objects I want to construct a simple test of a 3d rotating cube.

That way we will get a solid base established and then we can think about how to implement the other objects. It will also give the developers of such objects a way to test their objects to see if they work. I plan to make the interface public here so that any developer can tackle a certain area of Ogre (Although the plan is to go ahead and map out all of the objects and then developers can volunteer to work on a particular one).

I will create these three objects myself and get them working. Then other developers can help tackle things like lights, matrices, vectors, skydomes, particle effects, terrains, GUI components, ect.

Guys, you have read this post haven't you?
http://www.clickteam.com/epicenter/ubbthreads.php?ubb=showflat&Number=45593&page=1#Po st45593

Cmon, don't distract vortex2, he's got the physics extension to finish and a pretty tight deadline at that.

Indeed. The Physics extension is just about done (for version 1.0 that is). I have a few things to finish up and then I will be releasing a beta test of it.

Can't wait Vortex. The 3D object is going to be a good one, but whether you realise or not the physics object is going to be a revolution for 2D games once it's out in the wild! You are going to find yourself becoming very popular and answering a lot of questions about how to get the most out of it :D

As for the 3D object, you sound like you have everything in hand. Can I just throw in some advice to consider as you continue to design it. Even though it is an MMF extension, you could get some ideas of how to implement an interface for it by looking at how other development systems have done it, eg: Jamagic and Blitz for example.

As a user, I'd be more interested in being able to create a world, add camera's and lights, import objects (static and dynamic), being able to move the dynamic objects (be they meshes, lights, cameras) in both the usual ways and with helper actions/conditions/expressions that do the simple and obvious stuff, being able to state how a dynamic object reacts with static and with other dynamic objects for collisions, being able to define animation sequences and play those back. If the things Ogre provides in it's SDK map directly to these principles then great, but as a user I probably would not be too interested in having a direct 1 to 1 mapping of every function, but more interested in being able to have full control of the elements that are important to making a 3D game or other 3D type toy. So for me it will be how successfully you are able to make an interfeace that compliments MMF as much as the underlying 3D technology.

I'm very pleased you have decided to lead in creating this object as I know you'll do a good job on it, along with everyone else you pull on board. But right now I'm more excited about the physics extension as that will significantly change the type of games we're going to see people making with MMF :)

Thank you :). I have not forgotten about the physics extension ;) .

I do agree with your concerns that the 3d extension needs to strike a balance between power and ease of use. It needs to be flexible and yet it should also do some of the complicated stuff for you.

The key and the hardest part is to strike an acceptable balance.

One thing though is that whether it is Fusion Dynamics or a 3d extension, they both represent a new breed of complicated extensions. The users of such extensions will have to deal with added complexity and while I expect the physics extension to go over well, I am fully prepared to hear a lot of complaining about how hard it is to use and how it doesn't have built in support for "Feature X Y Z" ;) . I think some users will love it and some will ask for me to include built in racing mechanics :P .

In a way power is better in my mind then straight up ease of use. I can always explain how to do something but it is hard to do something that the extension isn't capable of doing. I would rather have to explain a complicated extension then to have too weak of one ;) .

I agree :D