How goes Xith3D?

Hey guys… it has been a while I know since I last stopped by. I continue to be amazed by the interest and loyalty to Xith3D from the java gaming community. Some projects have cleared up for me and I am looking forward to pulling down the latest code and trying it out. I may have written the original version and the first year of enhancements, but the real work has been done by all of you. Without your work this would have been one more API littering the cutting room floor. I hope no one bears any resentment to my retirement from the project, life took a turn and I had to change my focus for a long while.

So tell me, how have things been around here? Any idea on an estimate on how many people are using the libraries? Did many companies ever adopt it for their commercial usages?

What are the current weaknesses of Xith3D in comparison with other java 3D libraries? Did anyone ever add shader support, volumentric fog, bezier patches? Was there ever a decision to add physics, springs, etc?

Has the evolution been toward low level optimization / card feature support, or higher level architectural additions?

Anyway… just curious… and it is great to see familar names around here!

woah, blast from the past :slight_smile:

It’s good to see you back David, are you going to become a regular around these parts again?

Good to see you back

Endolf

I had been wondering where you went:
http://www.java-gaming.org/forums/index.php?topic=3267.msg87491#msg87491

I did a Google search on your name at one point. Out of curiosity, what have you been up to?

After working on Magicosm for 5 years, and developing Xith3D along the way we hit a crisis point where we ran out of money, energy and people. It was the first big failure of my life (Magicosm) and it was a huge blow to me when it failed. I had persevered for a long time after other similar projects failed and had never really accepted the possibility of failure. We got all the way to beta test with about 15 people and finally realized that even with all the technology finished and working we were never going to be ready for production, not without a staff of paid modellers, artists, voice people and writers. Even the production costs like hardware, bandwidth, customer support, etc was going to be beyond us. We had always thought a grass roots, word of mouth product could boostrap itself into existance. About this time World of Warcraft came out, Guildwars and EverQuest II. All you have to do is spent 10 min in WoW to realize how much more that world had to offer from what we were building… and dispair set in. I retired from Magicosm and in an effort to deal with the pain of that, also pulled myself out of the whole world of game programming. I will never attempt such an ambitious effort again.

But I still love 3D, and the wonderful creative process of turning math into something visually appealing. So after a couple of years of distance between myself and Magicosm I am testing the waters again.

Dave

Great to see you back man! :slight_smile: Looking forward to seeing what (smaller scale) creations you come up with :slight_smile:

I do miss that minotaur though - was alot of fun bashing the hell out of him :slight_smile:

Kev

hi, David, welcome back.

Well, we don’t know each others personally, but you’re a (good) ghost always being namely present in the code. There’re still the most classes in the source originally authored by you. Well, if it is a good sign after all this time, is another question :wink:

Marvin

Dude!!! glad to see your name again…man its been a long time. So sorry that magicosm fell by the wayside. Your reapperance ruins the title of my gaming effort “Mysterious Mysteries: The Search for David Yazel”

Where are we?? Interesting moment in time here. I guess the term is synchrounicity, when seeming unrelated events meet in time. Many people have entered and left, we have some relatively new people to the group who have taken on release, UI, and API type work. Some of the changes have been happening rapidly with potential API regressions.

In general I think it is safe to say we have this set of tools that nobody really seems to know enough about. There all kinds of stuff scattered in the system from perlin, to shadow, to terrain systems that nobody is using 'cause there are no examples. Some people like Croft are using it professionaly.

Quick summary.

Physics: can be covered by Odejava, william works on this and it is a java layer on top off a C layer with Xith bindings. I currently use the inital Xith collision system with a back dated Xith release, cause the collisaion system died somewhere along the way.

Multipass rendering: Has been implemented

Loaders:  Support for Collada and the Goole skp formats was added.

Shader support:  A guy named javacooldude implemented some of this in his exxamples off the xith.org site.  We also have recent work [url]http://www.java-gaming.org/forums/index.php?topic=14297.0[/url]

UI: I am still using the UI system you wrote. There have been several attempts to redo this since it seems to be time expensive. Some current work is at http://www.java-gaming.org/forums/index.php?topic=14099.0

Performance: At one point we were faster than Java3D, recent Java3D work has optimized their code to run faster.

Hi ,
its good to have you again (hoping that this is the case ;))…
Welcome on board and about the question ,how many people are using it and what all purpose for is question i am also intrested in.

Hi, Dave, great to see you again… I think when I first came to these forums you were already gone. So really it’s good, good to see you here there are so many things to clear out about the core and other things.

I didn’t read hawkwind post, will do it after having posted so you can have two seperate point of view. I’m not the greatest contributor but heh I followed closely the forums (maybe to closely).

Thanks. Could you be a bit clearer about your present situation : are you willing to possibly spend some time on Xith ?
And let me correct you : the “real work” couldn’t have been done without your original code. Thanks, thanks and thanks again.

Pretty good.

I’d say about 30 but it’s just an imprecise estimate cause I’m sure many more use it silently (no sign of life on the boards).

I know JProof (where YVG is/was working) once used it but I don’t know if it still does so. I know a guy use it for a small commercial game of plane models simulation. Croft seems to be paid to work on Xith3D (it’s not clear really for me who is its employer).

3-4 people said it was “messy”, something we’re fighting against. But on this point some explanations from you would be very very nice.
Someones feels some features are lacking but the problem is rather that many features are not used yet present. But often code is half-done and not really useable. See terrain code (it’s yours isn’t it ?) it’s really really cool the demo is impressive but heh let’s say I’m the 3-message-noob-coming-from-nowhere and I want to texture it… How do I do ? Probably there are texture generators classes (I remember having seen something like that) but it’s not documented. If you give me some notes I’ll happily add that to the guide which is currently written by Qudus and me.
Some recent (3-4 months IIRC) benchmarks showed that Xith3D was really slower than jME and even than Java3D ! They just ported your Quake 3 Level loader for benchmarking. Do you have any idea of what can be improved/optimized ?

Assembly shader support was added by Abdul Bezrati long ago for its demos. High-level shader support was added by Florent Hoffman (Goliat) recently (just look at the recent threads).
No news about volumetric fog.
Bezier patches and splines were discussed by Croft. Maybe he’s gonna implement it someday (cause Collada include them).

I haven’t seen any low level optimization, probably cause the core is hard to understand.
Some important features were added : JSR-231 support (Lilian Chamontin), COLLADA loader (David Wallace Croft), a HUD system (Marvin Froehlich aka Qudus and myself), a new picking method (Arne Müller) and some others that I can’t remember just here.
Also Qudus and myself have made some attempt of higher level architectural additions. When the “Xith in a Nutshell” guide comes out (not it’s not published by O’Reilly editions, sorry) you’ll see how easy it has become to set up simple environments.
Also, mostly Qudus and also myself have made lots and lots of clean-ups (great thanks to IBM and the Eclipse team).

One more time great to see you too.

Just about physics : you may want to take a look at JOODE : http://joode.sourceforge.net/index.html
I followed it a bit (when I still had projects of mini-games making heavy use of physic simulation) and they’re using Xith3D for their development (visual debugging). Currently only “bindings” for Xith exists.

Hi David,

It is great to see you back! I use Xith3d in my flight simulator and I am one happy user. The lack of shades (or the existence of broken shades) somewhat bugs me. I am out of my league in Xith and can’t really fix it. Will you please take a look at the shadows and see if you can make them work again. It will be just awesome if you do that. And, again, as Xith3d user, I am happy to wellcome you back !

Stefan
http://rcflightsim.com

Hi David, nice to hear from you again in these boards !

I hope you will stay here for a while and contribute sometimes !

(if not, it would be kind to help us understanding the unused parts of xith in order to finish and document them)

Lilian

Hi David,

Great to hear from you, I hope life has been treating you well.

I ended up working on a project named Lineriders with Philip Worthington that used Xith3D and Odejava. It was a fun project and was part of a show last year in London at the Royal College of Art. It also helped me get a job in the games industry where I am now working.

I havn’t been active on Xith3D recently due to work taking most of my attention, but still lurk in the forum from time to time. It’s certainly great that there are still people (some new, some old) putting energy into the project and taking it forward.

Cheers,
Will