There are a few Examples from Ardor3D that I think look interesting. And I wounder if there’s some why to do the same things in JME3. I’m not counting implementing it from scratch since that would clearly be possible and how to do it generally would be available in the ardor source
Generated Textures (Page 2). I’ve seen the SDK plugin for hooking up different texture generators so I guess I could look at that. Any other ways?
Particle Swarm (Page 5). From how I understand JME Particle system, particles can’t be influenced after creation by anything but gravity. Or is there a recommended way to build external influencers? The Influencers that are there is for initial velocity as I understand it. Dock for getParticles say “Returns a list of all particles (shouldn’t be used in most cases).” Is building a swarming particle system a case where you would use it?
Quad Imposters (Page 7) I don’t find any reference to the work “Imposter” in the source but It’s possible it’s there under a different name. (or that the base technique is there and it’s “trivial” to built this on top?
Trail mesh: It’s not up on the site but it’s in the examples. It show’s a glowing trail after a moving point. I think it’s just a mesh that gets generated as the point moves.
I was unsure in what category to post this question. It's not really troubleshooting, and not a feature request as such. So I just posted it here in general, I hope that's ok.
Preemptive thanks to anyone who takes the time to answer my curiosities..
And then I can just take the algorithms for the textures from ardor (or other sources). Makes sense.
i guess the influencer system can still be improved, its relatively new, else just use a geometry with a custom control
OK, but you’re not aware if anyone already wrote a “swarm” control that could just be plugged in.
And regarding to the influencer system. I guess it could be improved. Parhaps “attractors” “deflectors” “bounds” and all sorts of things. But on the other hand it’s sometimes a good thing that the api has a lower comlexity so it’s easier to learn.
BillboardControl
That will make them face the camera yes. But what do would I use to generate them in the first place.
I’m new but I’m thinking probably something simular to how the shadow maps are generated? Just render the object to a texture when it “get to far away” and then replace it with a billboard quad with that texture.
a particle emitter set to “world emit”
But the “trail” is connected like a tail. it’s not “particle” based. But I guess I could adapt the particle system to it but my question was if parhaps someone already did that.
Here’s how it looks:
I think that sort of effect could be cool for things like sword “swooshes”, long connected rocket tails (parhaps together with some standard particles as well)
One of the reasons I asked is that when I was researching what engine to use I naturally ran the examples in both. And some of these made me feel that there as a lot “there” in ardor. So I found the swarm thing and the trail thing kind of inspiring. Also the projected grid water was really cool.
But there’s nothing stopping me from trying to port some of those examples to JME for the next person. I just feel I need to learn a bit more first. And if there’s work already done in a specific area It seems my time could be better spent elsewhere.
Ardor3D is based on jME2 and makes using shaders more complicated than jME3, almost all your needs are best taken care of directly at shader level so I guess jME3 is still the better solution for you.
2) Yeah, jME2’s particle API was very bloated and as its the same guys I guess the ardor one might be too (which on the other hand obviously gives you a lot of options in the first place).
3) You’d have to generate them yourself yeah E.g. a pre-view where you only have the model attached and can rotate and “photograph” it would be best for that.
@kemi said:
4) a particle emitter set to “world emit”
But the "trail" is connected like a tail. it's not "particle" based. But I guess I could adapt the particle system to it but my question was if parhaps someone already did that.
Here's how it looks:
I think that sort of effect could be cool for things like sword "swooshes", long connected rocket tails (parhaps together with some standard particles as well)
@thetoucher once had a video that showed an effect almost exactly like this... and some other similar things he did by accident. Maybe he could explain how he did it.
These things aren't "built in" to JME but you can still do them.
@pspeed said: @thetoucher once had a video that showed an effect almost exactly like this... and some other similar things he did by accident. Maybe he could explain how he did it.
These things aren't "built in" to JME but you can still do them.
That would be cool. And since JME is open source I can do "anything" it's just a mater of how much "I" need to do, how much is already done and how well it "fits" into the engine.
But yea. I've decided to use Jme now.
The things that makes me think it's the better option for me is:
- The shader thing you talked about. It seems more "natural" in JME. And most "fancy graphics" stuff is done at the shader level now anyway.
- The community. There's some activity on Ardor forums as well but nothing compared to here.
- The documentation. Both project could use some more but there's more of it in JME and the forums in JME really makes up for a lot of what's lacking.
- The "Complete Stack" approach. In Ardor there's no sound, no networking no physics etc. So even if I later deiced that I want do do networking different or whatever at least I can do the prototypes etc. without having to integrate all those things.
- The post process water is awesome :D
Some things that made me consider ardor more was:
- The examples looked cool.
- ArdorCraft API looked cool (But there's mythruna as well I've seen now and I'm not going to make a Minecraft clone at this time anyway)
- The code seemed more "organized". There's the maven stuff and mostly in JME there's many things in "Core" that I would not consider "core" (water and particles for instance) so it's hard to see what depends on what. (note: this is to my uninitiated eyes so It's quite possible it's better that way anyway)
- JME felt a bit bloated. I'm particularly not sold on the SDK yet.
@kemi said:
- JME felt a bit bloated. I'm particularly not sold on the SDK yet.
I'm sold, having everything linked together so that you can download the package do a wizard to setup a project, and click run and it works, that is super nice...
i think this is sure possible with particle emitter, but with lower fps there could be problems(with emitting particles too far), so better to use perfect solution
(and without making too much of a product placement, i think the next version of ardor3d will cover all your arguments against it. still awesome that we can exchange ideas, code and so on to keep the java+3d industry going in general! )
Most of these things we implemented already back in the jme2 days and should still be in there as well. Like the TrailMesh for example. (http://www.jmonkeyengine.com/doc/com/jmex/effects/TrailMesh.html)
Right. I'll keep that in mind when/if I get around to wanting these particular features. Thanks.
(and without making too much of a product placement, i think the next version of ardor3d will cover all your arguments against it. still awesome that we can exchange ideas, code and so on to keep the java+3d industry going in general! )
Are there big changes coming? Things not in svn trunk? If so: cool. I think it's a good thing to have alternative ways to doing things. I still see good things in both engines. But I think for now I'm going to stick with JME3 for the above mentioned reasons as well as: I just need to pick one and get on with learning to make games instead. The stuff I learn in jme should translate pretty well should I choose to switch to ardor later.
E.g. a pre-view where you only have the model attached and can rotate and “photograph” it would be best for that.