SCIJump releases are available here.

This page also has instructions on how to check out the software from Subversion.

SCIJump is not SCIRun

SCIJump is not SCIRun. SCIJump, however, does currently depend on many of the original SCIRun dataflow classes. Because of these dependencies (and for other historical reasons), building SCIJump also requires building SCIRun. Eventually, the original dataflow classes will be isolated from SCIJump and built only as an optional component.

Development status

SCIJump is currently in the alpha release stage.
The latest version is 0.1.1.

Supported platform(s)

SCIJump code is primarily built on Linux (RedHat, SuSE, Debian) and Mac OS X (10.4). In theory, it should be possible to configure SCIJump on any Unix-like platform.

Building SCIJump

Install from build script.

Build SCIJump from source.

Bugs

Bugzilla bug list

Reporting bugs

Report SCIJump bugs under the CCA product category.

Running SCIJump Software

Executables

Start SCIJump from the command line. Specify gui as the builder option for the graphical interface. Optionally load a previously saved component network input file net_file.

scijump [-]-b[uilder] gui/txt network_file

Parallel components

Executing parallel components in SCIJump requires that an MPI service be running.
If you're planning on running parallel components, start lamd.

(parallel component use through the GUI isn't working properly)

Parallel Components in SCIJump are instantiated on ploader resources. The process of
instantiating a parallel component is similar to that of instantiating distributed components.
The only difference is that the ploader should be started in parallel:

mpirun -np <num_of_parallel_processes> ploader <ploader_name> <frameworkURL>

How to create and use components in SCIJump

CCA Components

Getting started with Hello World.

A component authoring wizard is available for CCA components.

Babel Components

(coming soon).