I confess I had forgotten about that bug.

But the bugs aren't entirely contradictory:


id=35302&atid=414613> bug 1403220 says that the behavior of the ':'
function when one of the endpoints evaluates to NULL should be to return
the empty set; and that Richard should create a testcase for that
behavior, whereas

id=35302&atid=414613> 1671613
id=35302&atid=414613> says that the ':' should allow the NULL literal
as its end-points (plus some stuff about semantics which is inconsistent
with 1403220).

Let's keep the semantics of ':' the same, as stated in 1403220, for now.



From: mondrian-bounces (AT) pentaho (DOT) org [mailto:mondrian-bounces (AT) pentaho (DOT) org]
On Behalf Of michael bienstein
Sent: Thursday, March 01, 2007 3:13 AM
To: Mondrian developer mailing list; mondrian-devel
Subject: [Mondrian] Recap - that bug is already there and opposite


Look at bug 1403220. The decision was to keep current Mondrian
behaviour and not align with MS behaviour. Maybe a property for this?