I'm a .NET developer who is semi-educated in 3D software (MAX, Maya), and I spend a lot of time studying/doing game design, everything from card games to shooters. At this point in time I don't so much want to make a specific game, as just learn more about the 3D aspects of design. What's possible, what's not, etc. Still, I'd like to to make the program the prettiest pictures possible while doing it.
Anyways, that's me. Hi. Now I have a couple of issues spinning around in my head as I'm trying to figure out wether I should learn JME or somthing else (if there is a simple article or forum post answering them, feel free to just post a link to it, but I haven't found them). If anybody has some time and is interested, I'd sure like to hear some opinions / answers.
1.) Per pixel shading - The 3D artist friend of mine who'd be doing a lot of the art for me refuses to touch something that won't do this. The feature list shows Fragment Shaders, but most of the demos seem not to use them. The "bump mapped wall" demo shows this, but it says you have to load another program. Is this 3rd party? Is it stable? Is there a tutorial for it? How hard is it to do?
2.) Comparison to other engines. Specifically, Truevision 3D. I'm fine with using C# (which TV3D will let me do), as I know it better than Java. Have any of you used it? I've been reading about both, and it seems that TV3D has more features. They're also not as open, I'm having trouble finding any tutorials that are as in depth as the ones here, and I just feel "fuzzier" about what it actually does (the JME guys seem to say/show what they give you straight up). Most of the things I read about TV3D were from people biased twoards it one way or another. I'd be interested to know if anybody tried it and decided not to use it/what their reasons were. One thing I will note is that this community seems waaaay more friendly, responsive, and generally helpful. A lot of flames on the TV3D boards.
The apparent pros of TV3D are: more features (I think), that the demo screenshots look a little nicer (just my opinion), and that I happen to like C#. The cons are: I feel "iffy" about it, probably because I haven't found many cons. If one of you could please drag it through the mud for me, that'd be good.
I've done most of the JME demos, found them easy to do, easy to understand, thought that they yeided good results, and just generally liked them. I see a lot of pros to JME, but won't list them here since you already know them. Also, the community seems to be great: I asked for help with a tutorial a couple days ago, went upstairs to make some toast, and when I came back not only had the solution to my problem been posted, but the tutorial was corrected! Thanks!
So… yeah. That's what's buzzing around my brain. If you'd like to give me your opinion, I'd much appreciate it.
I can't say that jME is for you, but I see that TV3D set a water mark for the free version. It's one of the reasons I prefer free engines to semi-free engines like agentfx.
TV3D is great, it even works from Java (trough a COM wrapper). If you are satisfied with what features it offers. But you cannot touch the engine, no sources.
Hi, and welcome!
- The “demos” are meant to show only the thing they are demoing, and are not made to be pretty “selling the engine” programs. They work more like visual unit-tests or tutorials if you wish.
jME supports glsl shaders which can do per pixel shading. But “with great powers comes great responsabilities” s using shaders makes alot of stuff harder on the developer. different number of lights or special multitexturing methods means different shaderprograms, instead of just add/removelight, set the multitexturing options, like you get for free with the fixed function pipeline…
we have a few glsl based tests in the cvs. have a look at the things under the jmetest.effects.glsl and jmetest.effects.water
here’s something half-pretty at least: http://www.jmonkeyengine.com/jmeforum/index.php?topic=2857.msg22261#msg22261
- i've never used tv3d, but some pros/cons could be:
- java+opengl=platform independance, jme runs on windows, linux, mac
- completely open source, looking through the engine source is a really good way to learn. also, if you want to extend the engine with your own features for a game, having the source is a huge pro
- as you noticed, a really great, friendly and helpful community!
- easy to use
- features, stability etc are improving very rapidly. everyone can contribute to the source + several companies now uses jme profesionally
- moving over from c# to java is just a matter of changing the methodnames to start with lowercase (yeah it's a java ripoff from scratch flame flame)
- jme is still a young engine, hence the pretty basic featurelist (but we'll get there)
- we don't have any really nice demos that combines the more advanced features to "sell" the engine
- since it's opengl we don't get the nice utilities directx offers(d3dx etc)
lots more to say, but it's very early in the morning, and i'm up with my 1 year old daughter…too tired to think, so i appologize for not making any sence in anything i just wrote
in any case, i truly hope you decide to stay, and help taking this engine + family(community) to the skies!
Plus with Java, you get Java Web Start and web applet functionality, which AFAIK you don't get with .Net.
Hi all, thanks for the replies.
Yeah, I'm still a newbie with delusions of granduer, so I think I'll stick in the magical world friendly communities and open-sourcery that is JME, I think it will be best especially for learning and experimenting, etc, etc.
Sorry I took so long to respond. School started, I moved to a new place, and… in short, I've been really, really busy lately. So, this may be late, but I still wanted to say thanks.