Date: Thu, 3 Feb 2000 21:28:43 -0500 From: Bill Dorland To: gs2@kendall.umd.edu Subject: GS2 bug fixes Reply-to: bdorland@ipr.umd.edu A couple of GS2 problems have been fixed and posted. The first problem has plagued several of you, and was associated with having to fiddle with ntheta in some cases, particularly when using equilibrium_option='eik'. The affected files are: gridgen4.f90 rungridgen.f90 theta_grid_gridgen.f90 You should 'make' rungridgen and gs2 after getting these files from the website. I have tested the fix on the T3E. I expect that you will find that the problem is fixed on all other platforms, with no changes required for your existing input files. However, in the unlikely event that you have a case that still shows this bug, you will find that a new warning message is written to your screen, suggesting that it would be wise to stop the run. There is a new input variable that you can specify in the theta_grid_gridgen_knobs namelist. It is the real variable 'bpknob'. The default value is 1.e-8. The appropriate thing to do would be to try the run again with slightly larger values of bpknob. I suggest choosing 1.e-10 < bpknob < 1.e-6. The second problem was associated with getting a clean make on the first try. The affected files are Makefile (which is the same as Makefile.auto). Thanks for the feedback. I'm sorry it took a while to get the first problem above fixed. --Bill