Hi

Excel is behaving strangely browsing my virtual cube with Pivot tables.
What happens is that certain measures become blank in some
circumstances.

I'm completely familiar with the reasons why this _should_ happen: if,
for example, I use a dimension on the x,y or page axis that is _not_
defined in every cube in the virtual cube, then the measures from the
cubes where the dimension isn't used don't appear. This is by design,
and fair enough. What I'm talking about is something different - where
I use dimensions that are common between every cube in the virtual
cube. (I'll call these "common dimensions").

I have 13 measures defined in the cube. Depending on the order in
which I drag them into the pivot-table, different strange things
happen:

a) If I drag all measures into the data area, with no dimensions on
columns, rows or page, everything's fine.
b) If I then drag a common dimension onto the x,y or page area,
everything's fine.
c) But if I start off with measures from one cube shown, and a common
dimension, and then try to add in measures from another cube, one of
two things happens:
- The measures appear in the Data area, but only the measures
from one cube are actually populated - the other ones are blank, or
- Excel gives me an error "The field you are moving cannot be
placed in that PivotTable area".

I initially thought I'd designed the virtual cube badly. But browsing
the cube data in Analysis Manager, I found that that application is
perfectly happily displaying all 13 measures together, whatever I do on
the dimension axes (unless, of course, I start using a dimension that
only exists in some of the cubes - then the measures from the other
cubes disappear, as expected).

Anyone else come across this? Is this just done to "Excel pivot-tables
are a bit flaky as an OLAP browser"? I've found so little
documentation or NG activity on questions like these, that I wonder
whether people out there are using Excel PTs or have given up on them
in favour of a "proper" OLAP browsing tool.