>>>>Hi all,
>>>>
>>>>I just had a frustrating hour or more.
>>>>
>>>>A form with table buffering for specific table (of a .DBC), MULTILOCKS ON, etc...
>>>>
>>>>An APPEND BLANK, set the PK field, then SETFLDSTATE on PK field so it looks UNCHANGED.
>>>>In the Valid() of my subclassed TextBox I was comparing This.Value to the field's OLDVAL. Because it is a subclass the field can be anything so I deconstruct the ControlSource to construct the OLDVAL statement and I include the alias as a parameter.
>>>>
>>>>When the input field remained untouched (just tabbed through them) the compare never returned TRUE.
>>>>
>>>>I ultimately found that OLDVAL was returning .NULL. even though the field does NOT allow NULL (none of the fields in the database do).
>>>>
>>>>Is this normal? Checking OLDVAL in both VFP Help and Hacker's doesn't mention this.
>>>
>>>AFAIK SETFLDSTATE() does not work properly when used to make a row look "unchanged". HackFox3 says as much and I couldn't get it to work the way I wanted to in VFP5 either (don't know offhand if it was ever changed in VFP6).
>>>
>>>I have no idea if this is the root cause of what you're seeing.
>>
>>Hi Al,
>>
>>As I read Hacker's (6) the problem with SetFldState is that it cannot be used to prevent an update of the record. In my case I only use it to allow me to do a GETFLDSTATE later.
>>This part seems to work as I want. It was the .NULL. return that surprised me.
>
>Are you sure your parameter value is as expected, doesn't include trailing/leading spaces etc.?

Yes, that's why I spent over an hour tearing hair out - I checked every which way to Sunday to verify the values but never confirmed until desperate what OLDVAL was returning .NULL.