Jme-physics CVS-server down!


Like I readed from somewhere, the people from SourceForge were concerned about the slowlyness of the CVS-server of jmephysics. It has been very slow the whole day and now seems to be down!

I'm quite new to jme and jmephysics (my second day actually) and I was following the tutorial for jme & netbeans.

So, could someone helpful member please send me (somehow :wink: the actual cvs-package (files and the STRUCTURE) for jmephysics? Then I'm able to continue the tutorial…



Yes, darkfrog said he'd do that:

It's posted there now.  I just posted the JAR and natives but if the sourcecode is necessary to anyone I can post that as well. Let me know.


darkfrog said:

It's posted there now.

I'll try to get that up as well…my goal for this evening is to finally launch Roll-A-Rama after its long delay due to server issues and then a NAT bug with JGN I had to track down, so I don't know how much time I'll have tonight.  If nothing else I'll try to post it tomorrow morning…sometimes I forget things, so if you don't see another posting about this remind me…frogs are forgetful creatures. :-p


Yes, I readed your post about "Roll-A-Rama": networking with jme-physics, right?

I've been developing a tank game with a friendly guy from Sweden - using Darkbasic and Newton. The whole system is based on the idea that your own tank is "moved" by physic-engine and the other one is just "a hollow model". That way it's much more easier to use physics with multiplayer games. Are you using the same principle?

So you like frogs ;)? I prefer penquins. :wink:

So, please upload the whole source-package whenever you manage. I'll be waiting… :wink:

OLLI from Finland

Yes, I'm doing networking with jME-Physics but I'm using a different concept.  I'm doing synchronization messages that is an alternative to dead-reckoning to allow the objects to be synchronized as far as location, but in addition, synchronizing the forces currently acting on an object.  For example, if I have a beach ball that gets hit in the scene the server would send the synchronization information to the client of the current position as well as the forces and velocity currently acting on it.  This way, even if I completely lose connectivity that ball will be perfectly in sync unless a 3rd party (another client or something in the server) acts upon it.  The ball will perform on the client's physics world exactly as it should.  The clients and server each have a synchronized copy of the 3d physical world.


Sounds like the right thing to do, but does it really work that way? How often do you "syncronize" the coordinates and the vectors? The amount of "sended data" is not so much (you have to send x, y, z, xvelocity, yvelocity, zvelocity, xrotation, yrotation, zrotation), but I'm worried about the "lag" and the "corruption" = they start to go separate ways do they…?


Well, seeing as at any given point in time a message received synchronizes it to the most recent information I haven't seen many problems regarding even over 1 second lag times…sure, they are out of sync, but they're consistently out of sync. :wink:

Hopefully I'll have time to get Roll-A-Rama launched tonight and you'll be able to see it for yourself.  I'm so far behind my release schedule though we'll see if this actually comes to fruition. :-p


Sourceforge's CVS-server was up again so I got the sources, thank you!

So I don't need you to upload them but I don't know about anyone else…

Happy "Rolll-A-Ramaing!" :wink:

Cool…as usual, I've been delayed another day…hopefully this evening I'll launch.  The server is successfully runable now, which is a good sign. :wink: