New OS Support
OS X 10.8 (Mountain Lion)
SCIRun 4 will run on OS X 10.8 with XQuartz installed. If X is not available on your system when the SCIRun application is run, you will be prompted to download XQuartz. We are currently testing against XQuartz 2.7.2.
Subversion access information is available through our GForge project portal.
SCIRun 4 Installation manual
SCIRun installation manual.
Additional build notes
SCIRun source code can be found under the GForge project (the repository can be browsed from the project site).
The SCIRun build process is controlled through CMake. The build tree has been split up into three main pieces: Core libraries, SCIRun Dataflow (which includes modules, packages, and the scirun executable), and stand alone applications (such as Seg3D 1.x). SCIRun also now requires users to use the external dependencies that are distributed with the SCIRun source code. Building with the build.sh script on Unix-like platforms (see the Installation manual) will automatically configure and build these.
Note: Seg3D 1.x does not require all of SCIRun and (much smaller) binaries can be downloaded separately from the Seg3D page. Seg3D 2 has it's own repository.
Download patched this SCIRun 4.3 source archive or apply this patch to build SCIRun 4.x from source with a GCC 3.x compiler. Please note that we are no longer testing with GCC 3.x compilers, so we cannot guarantee backward compatibility in future releases.
If working from Subversion, change to SCIRun directory, and apply patch.
patch -p 0 -i SCIRun_4.x_gcc3.x-patch-20100303.patch
The SCIRun Datasets can be downloaded from the SCIRun download site. The compressed datasets are 390 MB in size. Individual datasets are also available from the NCRR Dataset Archive.
Data can be checked out using Subversion from the URL https://code.sci.utah.edu/svn/CIBCData/SCIRunData through developer access or anonymously. Browse the SCIRun Dataset repository from GForge.
Two major changes to the SCIRun network files have occurred since 1.24. First, we switched our nets from tcl-based .net files to XML-based .srn (SCIRun Network) files. You can still load .net files, but can only save in the .srn xml format. We strongly suggest you convert your .net files to the new .srn format.
We also changed the names of all of the modules (and many of the categories) to use a more consistent / intuitive naming scheme. This change renders old .net files non functional. However you can convert the old .net files to the new .srn format in a few steps.
- Copy your old .net file to .net.old so you have the original. The conversion script will overwrite your input file.
- Convert the .net module names with the provided script.
- Convert the .net to .srn with the scirun executable. The -c option loads an old tcl based .net file, writes the new xml based .srn file and exits.
example: to convert myoldnetwork.net :
cp myoldnetwork.net myoldnetork.net.old
python SCIRun/src/scripts/remap_modules_in_net.py myoldnetwork.net SCIRun/src/scripts/module-remapping.txt
./scirun -c myoldnetwork.net
Then you can run your newly converted .srn file like so:
Similarly for users who have been using SCIRun since the switch to .srn, there is a script for converting module names in any .srn files the new naming scheme.
- Copy the old .srn file to .srn.old. The script will overwrite the input file.
- Convert the .srn to the new module names with the provided script.
example: to convert myoldnetwork.srn :
cp myoldnetwork.srn myoldnetork.srn.old
python SCIRun/src/scripts/remap_modules_in_srn.py myoldnetwork.srn SCIRun/src/scripts/module-remapping.txt
This script can also run recursively on entire directory stuctures of .srn files by providing a path to the directory instead of to an individual file.
example: if you have all your nets in /user/joe/mynetworkfiles/ :
python SCIRun/src/scripts/remap_modules_in_srn.py /user/joe/mynetworkfiles SCIRun/src/scripts/module-remapping.txt
This page was last updated on 08/17/2012
Go back to Documentation:SCIRun