Thought you guys might get a kick out of seeing the other video we showed off at JavaOne, a time lapse of the NCsoft worldbuilder in action.
Got it up on my blog:
http://www.renanse.com/blog/2008/05/more-from-javaone.html
I must add that it's using only a very small subset of all the features of the tool. But still, if someone like me can create anything with it, we must have done something right s 
Man thats so sweet its almost dis-heartening 
(Can I ask how the texture adding is done?? is it just splatting?)
It looks very impressive! How long have you guys been working on the worldbuilder?
very cool, not bad at all for programmers art 
Sounds like this is an NCSoft product? Does that mean it have to be purchased or for internal use only or it will be donated to the community?
BTW, very neat product! 
Jeff
Sounds like this is an NCSoft product? Does that mean it have to be purchased or for internal use only or it will be donated to the community?
Internal use only, ATM.
We use a form of splatting, yeah. And the tool has been under development for much of the last two years.
That looks awesome. The music was great too.
Looks nice, only adds to my Pessimism in having NCSoft owning the brains of the devs.
Well, it wouldn't do a lot of good to contribute this tool to anyone… it's fundamentally tied to our asset system and to certain game concepts as well. Really, I see the relationship NCsoft has (if you can even call it that) with jME is as benefitial as you could possibly get. Here is a company that pays 3 jME devs (and 1 jME regular) to sit around and code with jME in a real world environment. Not only that, but NCsoft allows us to contribute back all changes, bug fixes and improvements we make to jME code. If they have us make a game and tools for that game, why should anyone feel entitled to get that code as well?
theprism said:
Looks nice, only adds to my Pessimism in having NCSoft owning the brains of the devs.
That's really nice of you. If I didn't work for NCsoft I probably wouldn't have time to mess with jME at all...
I completely agree with renanse's position and think that its great/wonderful that NCsoft allows the open source part of jME to flourish ~and~ that it pays for the development of jME. This works out favorably for both NCSoft and those of us out here in open-source land using jME for our own hobby and/or business oriented projects. Insofar as NCSoft proprietary tools built with jME go: it's their right to control their intellectual property as they see fit, afterall they paid for its development to suit their game design/purpose. Further on point, if anyone else out here progged a for-profit game with jME and jME built design tools, I doubt if other jME users would expect them to release their tool-chain code into the wild. For me, I take the video of the NCSoft world builder as a great example of what the rest of us can code toward and aspire to in progging our own world builders/editors with jME for our own projects and experiements.
The hunch behind is that three devs work for NCSoft, the effort being put back into JME at the moment from those devs is to generally improove the scenegraph with the specific NCSoft project in mind. Maybe if the devs were not tied to NCSoft they may have a broader approach to how the scenegraph is changing. Collada model loading for example presumably works for how NCSoft want it, but many users of JME seem to have real issues with it...
The tool per the video is exactly what JMonkeyWorld3d set out to do - why build your own - an asset system can just be an interface and an abstraction layer. The secrets should not be how to do something in JME, the secrets should be what the game is about.
It just seems so similar to an indie pop band that after signing up to a commercial record label change their tune to something that will sell more records and forget what they were passionate about and why they did music in the first place.
The remaining devs still seemed focused where possible, but think a breath of fresh air could be in order and letting in new blood to modernise the demos and perhaps to take advantage of threading ( DF made a good start but perhaps wasnt supported enough and have a full understanding of the opengl pipeline to fully exploit threading in the scenegraph ).
Renanse has already stated that on this forum JME is not a game engine, but the name is not changing to JMonkey or JMonkey Scene Graph. Its giving the impression it will be easy to write a game. Partly that is true if the game consists of a single TerainPage and a few funky effects, but it could be more and especially more helpful into guiding newer users into avoiding pitfalls. Think that a complete simple game from end to end would be very beneficial to the community.
Sun have recognised that the future of Java3d is grim and are on the edge of persuasion to a different graphics engine. A search on java.sun brings back few hits for JME, in my mind it could be alot bigger. Darkstar is on the horizon, is there anything complicating sun adopting JME cause its a hell of an oppurtunity??
Sure, its easy to feel hostile to what I have stated, but thats the way I see it going at the moment. Working ten hours a day on a game is very likely to bring you into a blinkered view. The future is yours...
I don’t think you understand. We have always worked on jME on our spare time (going up and down depending on how much time our jobs demand). The stuff that comes out from NCsoft is just a bonus, in addition to what we do on our spare time, like the ColladaImporter. If that was not created/committed from NCsoft there would probably not be any collada importer at all in jME. So us working at NCsoft (or Jadestone where i got to work on jME too) gives us the opportunity to do stuff for jME both on and off work.
I know that Mark/Josh looked at MonkeyWorld3D in the beginning (and so did we back at Jadestone), but it didnt fit our needs. And since it’s written against the Eclipse RCP (with SWT etc) it’s a pretty big commitment (we like swing and the freedom with that).
Our tool is heavliy tied into our “secret” game features. Almost everything it handles affects/depends on heavier systems we have built on top of jME, which offcourse can’t be shared.
In what way does the name jMonkey Engine give an impression that it’s a full game engine? It’s a 3d engine or graphics engine but with alot more features than just that.
Sun’s project wonderland has already gone over to jME from Java3D btw, and their devs, the old Java3D devs and lots of others (NASA etc) are going to help us take jME to the next level.
These kinds of posts just makes me wanna start up a new project which is all commercial and secret, where people don’t expect everything to be done for them for free. At least not without putting anything in themselves. Edit: ok i got carried away
Lets not get sentimental. When I post anything here that might help somebody (clue or even code) I'm just giving back the help I had from everybody. NCsoft is a company and needs profit, which comes from being more productive/efficient (whatever) and it's right to close source internal things (again, I don't think this would be a point to argue anyway).
And most of all I agree with something Renanse (and others) said. It's a bit naive to look for a "generic" game engine complete with tools, specially for free. There's no way one can build the "perfect" tool chain, integration and level editing tool. Can you imagine these scenarios:
1 - Need for speed carbon made with Cryengine;
2 - World of warcraft made with UnrealEngine;
There's one point I agree with the "pessimists": We really need more showcase demos… And a new website, and a lot more… JME is great but has its weak point we all know: needs better seamless animation support, doesn't run on consoles.
Knowing that, I think the best strategy (if there's such one) would be to make the community as big as possible, and for that, we need the things I just mentioned. Spreading the "piss" at JavaOne is great as everything done until now, so lets focus on some more uplifting suggestions (not only pats in the back).
Here's mine (not original, seen it for some time in the forum): Lets revamp the main site. Clean design, no fats… And with a SHOW-CASE showcase, with downloadable demos one can run straight from it (jnpl)… Everyone should be encouraged to create COMPLETE demos… The tests are nice, but are somehow difficult to try (first build with IDE, then get it running, then discover that hidden test package) for newcomers. I thought something in the line of XNA starter kits and demos…
I'm in to provide a racing showcase from the ground (physics, tracks, AI, audio and some FX - this would be my competing-demo) with (tutorial) documentation… Who's with me in this? We could make it a competition: "Best Game Kit"… Three months to make a demo like this downloadable (published in the same place? not necessary, but jnpl at least)… Then everybody can vote on the competitors (by running, playing and reading docs)…
Suggestion for required features:
- Main menu
- One playable level
- Pause/Resume/Exit/HUD
Better demos is something we really wanna get done. Awesome idea to have that as a competition (last one was way too long ago). I'll put one up
i agree with perick that it would be cool to have some more complete demos, thats why i started this post with the probably wrong topic name, but the idea behind it is basically the same
those demos need to be accessible somewhere, but not in the core or jmetest because it would bloat it up too much.
(This is all pretty off topic for a thread just showing off something we built in jME… but oh well)
I can understand that jME development might feel nc-centric at times, but honestly that is because the devs there have had the most time to work on the engine. Still, I believe most if not all of the functionality introduced from all the devs has been general in purpose, so please let's be fair here. I can't speak for the others, but I personally am still that guy who loves tinkering on jME, making it faster, trying to make it friendlier, etc. just as I was before I got hired on at a game company. I really want to see jME continue to be a shining star in the Java gaming community.
Regarding site and demos and such, I know these things are mentioned here often and the devs have talked about needing to improve these for a while as well. Unfortunately until recently it seemed no one had much time to make that happen. Heck, I also know that there are also many other things that would be nice… It would be great to have a more advanced animation system (we'd like that for our work projects too!) It would also be awesome to improve jME's importers and add support for new formats. I'd love to see a better foundational toolset… or a collection of tool pieces that could be fit together as makes sense for the game or app you happen to be making. We need better threading support.
I wish for all of these things.
Still, I think it's easy to suggest these things, and many people do, but as a friend of mine remarked, it's an open source project… and as such please don't complain when the magic code fairies don't deliver them to you. Don't get me wrong, there are many members in the community who have gone above and beyond to contribute code examples, wiki entries and small demos, etc. to help move the community forward. But we need even more code contributions if we want to grow the API in as many directions as we are all pushing. I think the devs have shown themselves to be responsive to contributions, especially those that have been made with easy integration into jME in mind. Where we've had code contributed that is ready made to slide into the api (the X3D importer is a great example) it has gone in quickly. Bugs pointed out here also get fixed quickly. If a contest will help spur more of these kinds of contributions, I'm all for it.
To help facilitate receiving more contributions, and help keep us on track and focused on jME's mission of being a professional quality Java scenegraph and gaming api, the devs are indeed working behind the scenes on a few things. Mojomonkey is working on a possible new website look that is more about showcasing the api. I've been arranging more collaboration with experienced 3d architects from groups like the old Java3D team and Project Wonderland, NASA, and other such folks at gaming studios using jME. As a dev team we've been discussing next steps and putting together ideas using suggestions from the forum community. Finally, we've also been putting together what might become a true architecture board to help review future direction and keep things generalized beyond any one team's needs or wants (still under development, please don't start nominating people or anything like that yet.)
With all of these things, I truly hope and believe that we are at the start of a very exciting and community benefiting development here with jME 2.0.
Wow, it really amazes me the attitude some of the users take towards NCsoft's involvement with jME. There's not a developer here that doesn't primarily drive the features going into jME from the project they are currently working on. The fact that NCsoft actually is paying for that development and chooses to give any of it back I think is a generous thing for NCsoft to do. They could have easily just forked jME for their own internal commercial development and there's no reason that it would be "unfair" to do so. However, they generously make contributions of the majority of the functionality back to jME.