[Apologies if this is distributed two times; my first attempt was from
a unsubscribed account.]
Dear NCL'ers,
The research group I'm affiliated with is using NCL to generate
2D plots of analyses and forecasts from the ARPS weather model. Our
system is set up so that an NCL script customized for a particular time,
domain, etc. is created by a Perl script each time the model is run.
After the NCL script is run and a NCGM file is produced, the NCGM file
is converted to GIF or PNG files for display on the Web. In the past
this approach has worked well, but recently we have been experiencing
segmentation faults after adding vertical cross-sections into the mix.
I have found that the segmentation faults go way when I cut out certain
chunks of code, but they do not go away when I put in hardwired
"if (False) then" statements that skip over the code. Given this
behavior and the fact that the problem scripts we produce are about
~19,000 lines long, my question is: Is there a limit on the number of
lines an NCL script can have? Could violating such a limit cause a
segmentation fault?
I'm also looking at other possible causes for the seg faults, but I
am running out of culprits.
Thanks,
-Eric
**************************************************************************
Eric M. Kemp E-Mail: ekemp AT ou.edu
Research Scientist Phone: (405) 325-7992
Center for Analysis and Prediction of Storms FAX: (405) 325-7614
The University of Oklahoma http://caps.ou.edu/~ekemp
**************************************************************************
_______________________________________________
ncl-talk mailing list
ncl-talk AT ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
This archive was generated by hypermail 2b29 : Tue Oct 29 2002 - 15:34:14 MST