- Introduced a command history log for statistical and

syntax checking input purposes
- Rectified an error message in fourmess.c
- HMcontrol did not check for the HM to stop before returning. This
  caused weird data files at AMOR as the data had not yet been downloaded
  from the HM.
- Fixed an issue about parameters in multicounter
- Temporary fix in nxscript.c to always read the Hm from the HM and not
  a buffer. This is prior to rethinking caching strategies for old style
  HM's.
- Synchronize now copies fixed motors correctly. This used to cause
  irritation with users. This now requires a script syncdrive to exist
  in the sync server which takes care of handling the fixed flag when
  this is desired.
- Added initify to sicsdata in order to copy large value timebins over
  properly at AMOR


SKIPPED:
	psi/amorstat.c
	psi/make_gen
	psi/makefile_linux
	psi/polterwrite.c
	psi/sinq.c
	psi/sinqhttp.c
	psi/sinqhttpprot.c
	psi/sps.c
	psi/tdchm.c

exebuf.c

@@ -18,6 +18,7 @@
 #include "dynstring.h"
 #include "exebuf.i"
 #include "status.h"
+#include "commandlog.h"
 
 /*-----------------------------------------------------------------------*/
 pExeBuf exeBufCreate(char *name)
@@ -216,6 +217,7 @@ int exeBufProcess(pExeBuf self, SicsInterp * pSics,
           /* print only SICS commands */
           SCPrintf(pCon, eLog, "%s:%d>> %s", self->name, self->lineno,
                    cmd);
+          WriteCommandHistory(cmd);
         } else {
           /* debugging */
           /* SCPrintf(pCon, eValue, "%s:%d>> %s",self->name,self->lineno,cmd); */
@@ -245,7 +247,10 @@ int exeBufProcess(pExeBuf self, SicsInterp * pSics,
     if (SCGetInterrupt(pCon) >= eAbortBatch) {
       SCWrite(pCon, "ERROR: batch processing interrupted", eError);
       SetStatus(eEager);
-      return 0;
+      if (pCall != NULL) {
+         InvokeCallBack(pCall, BATCHEND, self->name);
+      }
+     return 0;
     } else {
       SCSetInterrupt(pCon, eContinue);
     }