Difference between revisions of "SCIRun Scripts on SPL Machines"

From NAMIC Wiki
Jump to: navigation, search
Line 1: Line 1:
 
== High-Throughput SCIRun Scripts on SPL Machines ==
 
== High-Throughput SCIRun Scripts on SPL Machines ==
  
SCIRun scripts running SCIRun -E in bash were initially not working despite SCIRun building fine and all the manual nets working.  The script nets and manual nets are in theory the same except for paths etc.  Ultimately trashing the /SCIRun folder containing "on-the-fly-libs" and "tmp" in entirety and then deleting ".scrirunrc" and letting SCIRun regenerate both allowed the scripts to run.  Note this is not in the SCIRun svn directory such as  "/trunk" etc, but created the first time you run SCIRun, usually created in your ~/ directory.  This should not be an issue if you have never run SCIRun on a machine before, but might be if you have prior versions.
+
SCIRun scripts running SCIRun -E in bash were initially not working despite SCIRun building fine and all the manual nets working.  The script nets and manual nets are in theory the same except for paths etc.  Ultimately trashing the /SCIRun folder containing "on-the-fly-libs" and "tmp" in entirety and then deleting ".scrirunrc" and letting SCIRun regenerate both allowed the scripts to run.  Note thse files are not in the SCIRun svn directory such as  "/trunk/" etc, but created the first time you run SCIRun, usually in your ~/ directory unless you specify otherwise.  This should not be an issue if you have never run SCIRun on a machine before, but might be if you have ever run SCIRun prior on a given machine.
  
Simply deleting and regenerating "on-the-fly-libs" and ".scirunrc did" not work.  The "SCIRun folder", then ".scirunrc" needed to both be deleted and then SCIRun run.  I am not sure why the scripts are more sensitive to this but with some builds you might need to trash these to run SCIRun -E with scripts, despite it working just fine in the standard manual situation for which SCIRun is most typically used.  I have also not seen it on other machines. A hint would be if it tries to update ".scirunrc" everytime a new SCIRun is opened in the script.
+
Simply deleting and regenerating "on-the-fly-libs" and ".scirunrc did" not work.  The entire "SCIRun folder", then ".scirunrc" needed to both be deleted and then SCIRun run.  I am not sure why the scripts are more sensitive to this.   A hint to need to do this would be if it brings up screen every time it tries to update ".scirunrc" everytime a new SCIRun is opened in the script, or if your nets called by a script simply don't run, when the manual ones do.

Revision as of 10:16, 2 August 2007

Home < SCIRun Scripts on SPL Machines

High-Throughput SCIRun Scripts on SPL Machines

SCIRun scripts running SCIRun -E in bash were initially not working despite SCIRun building fine and all the manual nets working. The script nets and manual nets are in theory the same except for paths etc. Ultimately trashing the /SCIRun folder containing "on-the-fly-libs" and "tmp" in entirety and then deleting ".scrirunrc" and letting SCIRun regenerate both allowed the scripts to run. Note thse files are not in the SCIRun svn directory such as "/trunk/" etc, but created the first time you run SCIRun, usually in your ~/ directory unless you specify otherwise. This should not be an issue if you have never run SCIRun on a machine before, but might be if you have ever run SCIRun prior on a given machine.

Simply deleting and regenerating "on-the-fly-libs" and ".scirunrc did" not work. The entire "SCIRun folder", then ".scirunrc" needed to both be deleted and then SCIRun run. I am not sure why the scripts are more sensitive to this. A hint to need to do this would be if it brings up screen every time it tries to update ".scirunrc" everytime a new SCIRun is opened in the script, or if your nets called by a script simply don't run, when the manual ones do.