[Mondrian] NULL and EMPTY again
julianhyde at speakeasy.net
Thu Mar 1 20:36:39 EST 2007
First off, given what happened with Richard, I'd say that everything
I've said on NULL and EMPTY needs to be thoroughly tested on the MS
product. I get the feeling that NULL is never a valid value for scalars
according to some comments by Mosha. Yet there is probably some
property somewhere for NullHandling.
I couldn't agree more.
I repeat: If committers don't have access to an instance of MSAS to test
behavior against the 'reference implementation' of MDX, I can make it
available to them.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Mondrian