timucin!
As I have been reading the code, RFEXI cleans up all ongoing jobs (including REM&RER jobs), so howcome is it possible that this command does not actually (or really) at least try to start the switch? Please explain!! If you give RFEXI, ALL ongoing commands or printouts are "cleaned". Also ALL ongoing jobs are erased. So, RFEXI is not depending of RFxxx commands. Do you agree?
fcpli///
patcher: You are right, this has nothing to do with the original fault, but...
same