On 06/22/09 15:11, Regina Henschel wrote:
That makes it not clearer and I do not see by which rules OOo results in 702. Would you be so kind to give a step by step calculation?

And besides the question what OOo does, there is the question whether that conforms to the spec.

Calculation starts in (A7=)MULTIPLE.OPERATIONS(A5;A2;A6). This means, "Take the formula in A5, replace all occurrence of A2 with A6 and calculate the formula then". Right?

So instead of calculating (A5=)MULTIPLE.OPERATIONS(A3;A2;A4) OOo calculates "=MULTIPLE.OPERATIONS(A3;A6;A4)". And now? If I understand the spec correct, then OOo has to go to A3 and look whether it depends on A2 and replace A2 with A6 if so.

Right. A1+A2 is calculated as A1+A6, which is 2+700, or 702.

If the formula on which a cell depends is a simple operation like + or * (as in the examples in the spec), then it makes no difference, when the replacement is done. In addition, it makes no difference whether you use the reference or use already the value to which the reference points. But here it is more complex and need to be defined.

Another example, again enter in column A.
2
3
5
7
11
13
17
19
=OFFSET(A2;1;0)
=A9+A2
=MULTIPLE.OPERATIONS(A10;A2;A4)

Try the same with
=OFFSET(A2;1;0)+A2
in cell A10 instead of =A9+A2
I would expect the same result, but it isn't.

OFFSET doesn't mix well with MULTIPLE.OPERATIONS because it's one of the functions that cause the cell to be recalculated on any change, like RAND or NOW. So the changed formula depends on the special "ALWAYS" range, not on A2. That's a matter of implementation, and could probably be changed, but it doesn't look like a high-priority issue.

Niklas

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to