Hi Utkarsh,

The homebrew approach would be preferable, since configuration can be so 
tricky, particularly with Catalyst.  I don’t have to have a redistributable app 
bundle, just a working “linux-style” install that I can readily recompile as I 
modify our MPAS-O reader,  etc.

The failure is that “import paraview” fails and it is clear that python cannot 
find the library.  This makes me wonder if pvpython in the install 
directory was built correctly.  The pvpython in the bin directory does not have 
this problem, but the md5 sums between all pvpython built are inconsistent with 
the bin directory version the different, and working version, from those in the 
install directory.  

Do you know if this issue is just with incorrect intuition about how the 
directory structure should work or do you think I am building it incorrectly?

I’m also having trouble getting Catalyst to work properly because it is 
initially missing MPI4py (I installed this from source using pvpython, fixing 
that problem) and then fails when it calls pvpython with errors like

Traceback (most recent call last):
 File "<string>", line 2, in <module>
 File "mpas.py", line 187, in DoCoProcessing
   coprocessor.UpdateProducers(datadescription)
 File 
"/Users/pwolfram/src/ParaViewSuperbuild-gnu-make/paraview/src/paraview-build/lib/site-packages/paraview/coprocessing.py",
 line 103, in UpdateProducers
   self.CreatePipeline(datadescription)
 File "mpas.py", line 141, in CreatePipeline
   self.Pipeline = _CreatePipeline(self, datadescription)
 File "mpas.py", line 133, in _CreatePipeline
   class Pipeline:
 File "mpas.py", line 135, in Pipeline
   CreateMPASViews(pv_output_, variable_action[0][0], datadescription)
 File "mpas.py", line 350, in CreateMPASViews
   RenderView1.UseInteractiveRenderingForSceenshots = 0
 File 
"/Users/pwolfram/src/ParaViewSuperbuild-gnu-make/paraview/src/paraview-build/lib/site-packages/paraview/servermanager.py",
 line 295, in __setattr__
   "to add this attribute.")
AttributeError: Attribute UseInteractiveRenderingForSceenshots does not exist.  
This class does not allow addition of new attributes to avoid mistakes due to 
typos. Use add_at
his attribute.

when we utilize existing working scripts for our in-situ catalyst adaptor for 
MPAS-O.  Do you have any idea what could be the trouble?  I’m using the 
bleeding edge version of paraview, but the script was designed for 4.0.1.  

Any general ideas / feedback on these issues would be greatly appreciated.

Thank you very much,
Phillip 

On May 23, 2014, at 2:01 PM, Utkarsh Ayachit <[email protected]> 
wrote:

> Philip,
> 
> The superbuild is not really designed such that you can use the
> components it puts under  its internal "install" directory. That
> "install" location is indeed treated as a internal sandbox where it
> builds all the components it will finally installs.  The "make
> install" however should have installed a working version of pvpython.
> Can you elaborate more on what fails? Does it complain about a missing
> library?
> 
> Also, I thinking on OsX, we may be better of providing support for
> homebrew, and using the Superbuild only to generate redistributable
> app bundles. Does that sound like better approach, in your usecases?
> 
> Utkarsh
> 
> On Wed, May 21, 2014 at 11:53 AM, Wolfram, Phillip J <[email protected]> 
> wrote:
>> Dear ParaView Developers and Users,
>> 
>> After compiling ParaView via the SuperBuild process I have noticed that it
>> is possible to “import paraview” in pvpython in the
>> paraview/src/paraview-build/bin directory, but not in the
>> install/Applications/paraview.app/Contents directory.
>> 
>> I am on OS X 10.9 (Mavericks) using clang / gfortran via Homebrew and am
>> confused about this behavior because ultimately it appears that a “make
>> install” installs the broken pvpython in the install directory and not the
>> correct, working pvpython in the paraview-build directory.
>> 
>> This is strange because the binary installer which has pvpython in
>> /Applications/paraview.app/Contents/bin/pvpython works correctly but the
>> compiled from source installation does not.
>> 
>> Could someone provide me with some perspectives here?  I have included
>> CMakeCache.txt to provide more info...
>> 
>> Thanks,
>> 
>> Phil
>> 
>> --------------------------------------------------
>> Phillip Wolfram, Ph.D.
>> Postdoctoral Research Associate
>> T-3 Fluid Dynamics and Structural Mechanics
>> Climate, Ocean and Sea Ice Modeling
>> Los Alamos National Laboratory
>> Phone: (505) 667-7744
>> Email:   [email protected]
>> 
>> 
>> _______________________________________________
>> Powered by www.kitware.com
>> 
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>> 
>> Please keep messages on-topic and check the ParaView Wiki at:
>> http://paraview.org/Wiki/ParaView
>> 
>> Follow this link to subscribe/unsubscribe:
>> http://www.paraview.org/mailman/listinfo/paraview

_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at 
http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: 
http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview

Reply via email to