Tore,

yes this might be a reason. But the change is not documented (at least I didn't found). There are some ugly parts of code that did use that "feature" and I run in one of those. ::)

Agnes
>>Hi all,
>>
>>I found a problem (or behaviour change) from VFP 8 to 9. I don't realy need help on this. It's just for information and because I didn't find it mentioned elswhere.
>>
>>I have a view "viewA" that is based on an other view "viewB".
>>If I do
>>

>>close tables all
>>USE viewA NODATA
>>

>>
>>the result will be in VFP 8.0 (and in 7,6,5)
>>viewB is opend with data, viewA without
>>the result will be in VFP 9.0
>>viewB is opend without data, viewA without
>>
>>SET("EngineBehavior") returns 70
>>
>>Is this known? Is there a workaround that gives the way VFP8 works back?
>>
>>Agnes
>
>This is from memory, so please forgive if I remember wrong, but I think this was mentioned at a conference a few years back. As I remember it, it was said that the change was made to be consistent with the documentation, and to be more logical. If you ask for Nodata, you should get nodata. I don't think there are any ways to turn this new behavior off, because the old behavior was considered more or less as a bug.