Page 1 of 2

SoC: fluid : windows/Linux -2005/09/06

Posted: Thu Sep 08, 2005 4:16 am
by lguillaume
Hello, Nils update his branch :
fixed several solver issues
- rewrote geometry init (intersecting objects now possible)
- merged with 2005-09-05 version
- removed extern/elbeem diretory
- added solver in intern/elbeem
- renamed BKE_fluidsim.h to LBM_fluidsim.h
- removed old test code
Windows :
7zip version
zip version
Linux by Kidb can be found at http://blendertestbuilds.de/index.php?d ... 910/fluid/

Posted: Thu Sep 08, 2005 5:12 am
by ZanQdo
Thanks for the build, here is a .zip mirror for the ones who does´t have 7zip

http://www.graphicall.org/builds/builds ... Fluide.zip

Posted: Thu Sep 08, 2005 7:18 am
by kidb

Posted: Thu Sep 08, 2005 8:42 am
by JoOngle
Your fluid blender build doesn't work. It just doesn't start.
(the 32 bit version)

Posted: Thu Sep 08, 2005 8:54 am
by kidb
JoOngle wrote: Your fluid blender build doesn't work. It just doesn't start.
(the 32 bit version)
Hmmm, works for me.
downloaded, tar -xzf blender..., cd blender..., ./blender.
maybe you could try an strace ./blender and see if something is missing.

Posted: Thu Sep 08, 2005 9:08 am
by JoOngle
It seems to use an "older" libphyton file than mine... I use libpython2.4.so

but your compile wants 2.3....

Posted: Thu Sep 08, 2005 9:13 am
by kidb
JoOngle wrote: but your compile wants 2.3....
ok, just a moment, please

Posted: Thu Sep 08, 2005 11:11 am
by kidb
kidb wrote:
JoOngle wrote: but your compile wants 2.3....
ok, just a moment, please
./blender
Using Python version 2.4
http://blendertestbuilds.de/index.php?d ... 910/fluid/
http://blendertestbuilds.de/index.php?d ... 2.4.tar.gz

Posted: Thu Sep 08, 2005 3:43 pm
by lucky3
Intersecting fluid objects doesn't seem to work, or i can't get it to work properly. I made two intersecting spheres as fluid, but this doesn't polygonize as expected. Only one sphere looks good.
Edit: I get also a lot of crashes when Resolution > 110.

Posted: Thu Sep 08, 2005 4:34 pm
by JoOngle
thanks kidb - but it still doesn't work, now it wants:

libstdc++.so.6: cannot open shared object file: No such file or directory

I don't know what you compile with to make it want these extra things....
Normal linux blender versions work just fine both coming from n_t , gabio and
erwin...

Posted: Thu Sep 08, 2005 4:42 pm
by kidb
JoOngle wrote:I don't know what you compile with to make it want these extra things....
Normal linux blender versions work just fine both coming from n_t , gabio and erwin...
Even my Linux installation is a testing build, so is my life. =8-))
I do a build on my nightly cvs build machine. Starting
.
.
.
Now

Posted: Thu Sep 08, 2005 5:03 pm
by JoOngle
he he he .. yeah - me too. ;)

(Don't have access to the fluid source...hence why I'm not compiling fluid versions myself)

Posted: Thu Sep 08, 2005 5:21 pm
by kidb
JoOngle wrote:he he he .. yeah - me too. ;)

(Don't have access to the fluid source...hence why I'm not compiling fluid versions myself)
why not from cvs? from n_t's branch?
Or try the brand new blender-2.37a-linux-glibc2.3.5-i386.p4.libstd5.tar.gz 0 days old build, where p4 stands for python2.4 and libstd stands for libstdc++5.

Posted: Sat Sep 17, 2005 7:58 pm
by malefico
why not from cvs? from n_t's branch?
Sorry but how can I exactly access to it ? How is it called the module ? "n_t" ?
(It's funny I couldn't found any link to sources yet)

Regards

malefico.

Posted: Sat Sep 17, 2005 11:05 pm
by lguillaume
malefico wrote:
why not from cvs? from n_t's branch?
Sorry but how can I exactly access to it ? How is it called the module ? "n_t" ?
(It's funny I couldn't found any link to sources yet)

Regards

malefico.
Hello :
cvs -d:pserver:anonymous@cvs.blender.org:/cvsroot/soc-2005 login
cvs -z3 -d:pserver:anonymous@cvs.blender.org:/cvsroot/soc-2005 checkout -r name_of_branch soc-blender
cvs -d:pserver:anonymous@cvs.blender.org:/cvsroot/soc-2005 logout

Where name_of_branch is N_T for fluid, I didn't try I use tortoiseCVS on windows and it easier.

I found the doc at http://wiki.blender.org/bin/view.pl/Ble ... OfCode2005