Topic: ODE INTERNAL ERROR 2: Bad argument(s) in dxSphere()

[Recon]

Yep, downloaded Build 1004 and updated the other apps , ODE/CEL/cegui/CS and get this error now when i try run precursors.

ODE Rev 1048
CEL Rev 2183
CS  Rev 25244
CEGUI Rev 1390


error is

mec@mec-desktop:~/development/precursors$ ./precursors
preClient version 0.3.4 Build 1004 by G33X Nexus Entertainment.
WARNING: could not load plugin 'crystalspace.sound.render.software'
WARNING: could not load plugin 'crystalspace.sound.loader.multiplexer'
WARNING: could not load plugin 'crystalspace.modelconverter.multiplexer'
WARNING: could not load plugin 'crystalspace.mesh.crossbuilder'

ODE INTERNAL ERROR 2: Bad argument(s) in dxSphere()
Aborted

all seems to compile without any problems.
Only thing I have not updated is the gamedata, with the svn repository moved i not sure how to update it.

In general things going well, getting a pain in the head trying to figure out these transforms with T2O and O2T conversions. tying to move the camera around and make it look in the same direction as a mesh. Cant seem to get the angles right for the rotations lol. Been getting closer but no eureka moment when it starts to make sense    

regards

[Rosencrantz]

For now, try using revision 1021 of ODE (which is what I'm using and it still works). I'll see what I can verify and find as soon as things finish compiling.

EDIT: As far as migrating the game data, just cd into the gamedata directory and follow the instructions
listed here.

http://forums.g33xnexus.com/index.php/topic,589.0.html

EDIT2:

Also, currently working on getting an alternate testing box up so I don't keep overloading my main workstation when I compile CS.  :o

So it may be a bit before I verify this issue.

[Rosencrantz]

Recon, are you still having these issues?

[Recon]

Yep, tried to revert ODE and it did not help eather. Also the problem is showing up in other programs. Tried the wheeledtest in CEL and got same error. Cant be sure it's not just a problem with my system as no one else has reported it. 

So , Pritty much just waiting around for a update to fix it. Posted a message on the CEL forum so not much else i can do.

Regards

[whitelynx]

Check out CS's phystut program, and see if you run into it there. (that also uses ODE) If you do, try doing a 'jam distclean' on CS and rebuilding it. Also, triple-check and make sure you don't have any old versions of ODE sitting around. (check /usr/lib and /usr/local/lib for libode.a and libode.so, and /usr/include and /usr/local/include for ODE directories... if any of those are present)

[Rosencrantz]

You can also try out doing "lsof -p <pid of precursors>" and seeing where the libraries are that it's pulling. Good way to make sure you don't have any lost libraries floating about somewhere unexpected.

[Recon]

Well after getting on IRC with jorrit for a couple of min we fixed the problem. It was the ODE configure option "--enable-double-precision" causing the error.

Used
./configure --with-opcode --enable-debug --with-cylinder --prefix=$HOME/development/output

on ODE and recompiled , then recompiled CS, precursors is now running with out the DXSphere error. Think i will start a new tread for my next error

Regards

[whitelynx]

That's very strange... I always compile ODE with double precision... did Jorrit have any idea as to why that was causing a problem?

[Recon]

No, he just said that it was interesting and good to know!