Hobbyist or Professional?

[SiriusCG]

After reading a number of comments in another thread I started thinking about the Irrlicht Engine and what do people expect from it. For as long as I've been messing about with Irrlicht, I've always believed it was targeted towards hobbyist game devs.

So, here's the question: Should anyone using the Irrlicht Engine expect to realistically produce professional grade games?

Please understand, I, in no way mean to degrade or denigrate Irrlicht. It seems however, that a number of "professional grade" features are missing according to some and I'm wondering what ARE peoples expectations of Irrlicht now?

[kburkhart84]

I'm a hobbyist for now. I think Irrlicht as is could make professional games, but not $50 games. I'm thinking more like $10-$20 games. This is because of it's missing features. If some of these things were implemented as planned, then yeah, Irrlicht would b efine for high quality professional games as well. For the most part, Irrlicht is fine. You have to remember, Irrlicht is just a tool, a good tool for me, but a tool. The artists, designers, and musicians create the real game, whereas irrlicht just draws it. If you have good content and a fun game, then it should sell, even if it doesn't have the newest features just yet. If you have crappy art, design, etc... then don't expect fancy Irrlicht shaders to make it look good for you.

EDIT**This makes my 100th post. Cool.

[hybrid]

I'd rate Bolzplatz (not pure Irrlicht, though) and H-Craft Championship professional games. The latter will even go commercial soon. And extrapolating some demos to a full game could even lead to more pro-state games. So after all I guess it's possible and intended. We will continue to develop Irrlicht's features and usability, but in the end it's the gameplay and the artists' skills which will make the result a (commercial) success or just a hobbyist's demo.
So if you really find something's missing in Irrlicht and your desired effect is to no avail possible in Irrlicht just make a feature request and we will put it on our list (with working code it could be even faster to be merged). And if it's just speed then show us the slow results. If they are mind-blasting we will defintely work on improving the speed

[Nodtveidt]

Been a hobbyist game developer ever since retiring from the professional game dev world close to a decade ago. I'm unlikely to go back to being pro, although the current industry does give indies a chance to shine as professionals so it's still rather tempting...as far as Irrlicht goes, there's no reason why it couldn't be used in a professional game, it's just that people believe it "lacks something" that other commercial engines have, and since developers want all the latest and greatest features (although most of which they never use anyways), plus the fact that it's free ("you get what you pay for" mentality), Irrlicht gets left behind.

[SiriusCG]

it's just that people believe it "lacks something" that other commercial engines have, and since developers want all the latest and greatest features (although most of which they never use anyways), plus the fact that it's free ("you get what you pay for" mentality), Irrlicht gets left behind.

Well said and very much the point of this poll & thread...

[battlestar]

Irrlicht is the best suited open source engine for professional development mostly due to its no-strings licence, and it is a reasonable base for developing on next and current gen consoles too. The software renderers are good for Intel graphics chipset notebooks. As already mentioned above, for making a professional game one needs professional artwork. That's where most hobbyists fail. The engine features are far less important, look at GTA.

Wii too shows that you don't need every engine gimmick for a good game, and Irrlicht is fine for engine coders who can add features as and where needed. Shaders are supported, the rest is up to you. There are a few features missing here and there in Irrlicht (for example UV Tiling on textures). Not having an own import format for animations and meshes is also not good, in particular if the support for the different import formats is incomplete and it is down to the engine users to find out what works and what not. Why not adding unit tests to the importers?

It looks as if Irrlicht could benefit from more focus and planning of the development. The lack of a genuine roadmap document and directed proposals where to improve modules shows it, but correct me if I missed something there. Might be a good start to list the "professional features" the engine is lacking.

BTW, Irredit is nice for static scenes, but integrating Radiosity (even if slower than slow) into the engine would be excellent. Luke's pending improvements on the animation system are much needed and welcome too.

PS: Just waded through all the posts under "IrrEdid and Friends" after writing the above. Much has already been said, und much of it is true.

[hybrid]

UV Tiling is possible by using texture coords other than from [0,1]. What was not possible before was texture offset or easy access to texture coord scaling, but that's remedied by the texture coords which will be available per layer soon.
But yes, making heads up on the next large steps is always welcome. And making it in the public also enables the use of polls and further discussions on the overall architecture.

[battlestar]

Wouldn't it be good to set up a Irrlicht steering committee that identifies and communicates areas of the engine where improvement is needed?

Here some links and feature requests:

* feature requests
* really old requests
* bugs

some random but recent requests from the forums, partially obsolete, partially not. If stuff works for .x but not for .m3d then that's not really good enough. Ok, here the examples:

http://irrlicht.sourceforge.net/phpBB2/ ... c&start=90

A selection of limits I have encountered:
1. 4 textures - just about enough for my Doom3 models with diffuse/normal/gloss/displacement maps.
2. no NPOT rendertargets for DX - rendering an exact copy of the framebuffer not possible.
3. No per texture-stage settings - a single wrap for UVW in one texture stage and applying to all texture stages.
4. Single filter setting applies to all texture stages.
5. Rendering to multiple rendertargets during a scene - scene is updated at every smgr->drawAll() call leading to all rendertargets being out-of-sync. If you pause the timer, things like cpu skinning are still redone for every drawAll().
6. No HW buffers (DX: VB/IB) support - its quicker for me to render an entire Quake3 level in IrrSpintz than to use an octree to "optimise rendering" (about 4 times quicker).
7. tiny.x currently renders at 78fps for me. My MultiAnimation demo can render over 100 tiny.x faster than that and they're being blended between two animations.
8. Shader constants: float only. No bool or int.
9. Can't have a rendertarget bigger in width or height than the original window in DX, as depth buffer is shared.

omaremad, http://irrlicht.sourceforge.net/phpBB2/ ... c&start=45

* Another big issue with Irrlichts itself is leak of hardware vertext buffers
* the fact that irrlicht cant have a node that loads sevral textures and still be bumpmapped or parallax
mapped. eliminating dynamically lit levels.
* for example we cant add materials on a sub node basis, a whole node carries just one material, im sure many old games had the ability to have a transparent window on a car with cube mapping while still being under one object and not creating a node for it.

To bring the render ing quality up i had to:
-implement HDR
-On the fly parallax map genration with light map suppourt(this required invasie surgery into the sevral source files)
-dof
-softshadows(this required invasie surgery into the sevral source files)

[hybrid]

Bug reports and feature request should be filed in the trackers at the Irrlicht sourceforge project page. For a more concise and changeable listing one could also add new pages to the Irrlicht Wiki.

[SiriusCG]

Wouldn't it be good to set up a Irrlicht steering committee that identifies and communicates areas of the engine where improvement is needed?

Interesting idea...

[battlestar]

Bug reports and feature request should be filed in the trackers at the Irrlicht sourceforge project page. For a more concise and changeable listing one could also add new pages to the Irrlicht Wiki.

If you recommend it I'll enter them into the lists, even though they aren't really my requests but someone elses. Shall I?

best

[hybrid]

Soem time ago bitplane went through the forums to collect all those requests and bug reports. They weren't his either. So go ahead and add them. This will ensure that they won't get lost that easily.
For the improvements list the Wiki is better suited since these items will change often. Moreover we can easily move parts from the wish list to the todo list which is also listed in the Wiki (though read-only for non-Devs). But that way the dev team can easily see the current demands from the community while you can see what we're working on for the next release

[SiriusCG]

For the improvements list the Wiki is better suited since these items will change often. Moreover we can easily move parts from the wish list to the todo list which is also listed in the Wiki (though read-only for non-Devs). But that way the dev team can easily see the current demands from the community while you can see what we're working on for the next release

Wonderful!

[dwmitch]

It depends on your audience. That goes for any engine. If you're targeting the yokels who want pretty pictures, even if it means that the designers didn't have time to come up with a decent storyline and the control scheme is on par with the Atari 2600, then there's going to be at least one feature missing that will cause them to uninstall the demo after the scene loads (not sure what the feature(s) may be as I don't know Irrlicht's full feature list and I never progressed beyond DX8 in technological terms).

However, if you target the more intelligent people* who realize there's more to a game than graphics then everything from Game Maker (and similar programs) to Irrlicht to DX10 is capable of developing a professional program.

In fact, even though about the best you'll get out of Game Maker is SNES quality (it has 3D features, but they're laughable), there are forum members who do make a decent income off of it. One woman uses it to make games for adult websites, and while I haven't seem them (nor do I want to) and we've never discussed figures, I get the impression that for a time it was her primary source of income. There's another woman who creates educational games. The income from that combined with her Social Security benefits brings in almost as much income as before she retired.

There are people who still buy the Exile series (point and click RPG, no transition when moving from tile to tile (you just pop from one tile to the next), not even on par with the first Dragon Quest). I assume they still sell, because you can still order them. Normally even a shareware company will pull a product when it quits selling.

The point is, if you keep your target audience in mind it doesn't matter what tool you use. If you want the idiots who won't wait to find out that the game even has a story if they don't see HDRI lighting just as soon as the first scene loads then you might want to upgrade to Vista and use DirectX 10. If there's an online community obsessed with The Bard's Tale (PS2 version), then Irrlicht has the capabilities of doing a similar game graphically (though it won't handle any of the other elements, such as audio, etc.) and probably make a killing in that market.

If you want to target fans of the original Bard's Tale (google them both to see the differences if you're not familiar with them) then some of the abandoned tools from the 1980's and early '90's, assuming you can get them to run on the modern systems, would be professional grade.

Basically, there's no such thing as "professional grade" when it comes to graphics/physics/audio/etc. engines. Whatever you can use to make a game, assuming you can make a halfway decent profit on it, is "professional grade."

Of course, this only applies to the one man armies, the indie houses with no publisher, and the self publishers. When it comes to the big guys like Square-Enix and Midway there is such thing as professional grade, and Irrlicht is right on the border. It will only take one new major development in real-time graphics to push it back into the hobbyist realm.

*I know I insulted a lot of gamers, but I don't care. Ignoring a compelling story just because the graphics are DX8 quality shows a lack of intelligence.

[playerdark]

Well I certainly hope it is good enough for professional use at least I am using it for that purpose