>>Probably not the problem. I notice that there are 3 IF's and 4 ENDIF's.
>
>That was a cut and paste error.
>

What I thought.

>>
>>I have had problems with Buffering set to 5. Have you tried it at 4 or 3?
>
>Reason I used buffer mode 5 is I was trying to avoid locking the records untill the very last moment. How will changing buffering to 3 or 4 change that?
>

For testing I'd even turn buffering off.

>>
>>Which version of Novell? We are running 3.12.
>
>How do I determine the version of Novell we are using?
>

Your network admin should be able to tell you. You're probably not using 3.12, but if you are I would try to reproduce the behavior here. I haven't had problems with mapped drives - yet.

>>
>>>I have a problem which I cannot reproduce when I debug the version of the program on my C drive. If I copy the code to a mapped Novell drive I can reporduce the problem. The user runs a form and press on a button which invokes a "click" method. In it I check if tables where update and if not display an error message. I find that the second tapeudate reuturns ".F." Tables are buffered and buffering is set to 5 (optomistic table). Variables are declared as LOCAL variables. Billset is a cursor.
>>>

>>>llTableUpdateii	= TABLEUPDATE(.T.,.F.,'ii')
>>>IF llTableUpdateii		
>>>         *update issue records with this bill date
>>>         SELE billset
>>>	SET ORDER TO creqno IN pireq
>>>	SET RELATION TO creqno INTO pireq ADDITIVE
>>>	SCAN
>>>		llRlockpi	= RLOCK('pireq')
>>>		IF llRlockpi
>>>	                  SELECT pireq
>>>			REPLACE pireq.dbilled WITH m.idate
>>>		ELSE
>>>			llOk	= .F.
>>>			EXIT
>>>		ENDIF	
>>>	ENDSCAN			
>>>	         llTableUpdatepi	= TABLEUPDATE(.T.,.F.,'pireq')
>>>		IF NOT llTableUpdatepi
>>>		    llOk	= .F.
>>>		ENDIF
>>>		SET RELATION TO
>>>		USE IN billset
>>>		USE IN billtotl
>>>	ELSE
>>>		llOk	= .F.
>>>	ENDIF
>>>ENDIF
>>>

>>>I do NOT understand what is happening. Any help appreciated.