Issue 103291

Summary: MATCH and VLOOKUP in unordered mode fails with large numbers
Product: Calc Reporter: villeroy <villeroy>
Component: editingAssignee: AOO issues mailing list <issues>
Status: CONFIRMED --- QA Contact:
Severity: Trivial    
Priority: P3 CC: issues, jbf.faure, rb.henschel
Version: OOo 3.1   
Target Milestone: ---   
Hardware: Unknown   
OS: All   
Issue Type: DEFECT Latest Confirmation in: ---
Developer Difficulty: ---
Attachments:
Description Flags
Values and matches
none
wrong results for = and - with large integral values
none
The difference is right, when calculated via macro, but wrong when calculated directly. The error is not in IEEE but in OOo. none

Description villeroy 2009-07-02 19:43:08 UTC 
842020700000001	1	=MATCH($A1;$A$1:$A$5;0)
842020700000002	1	=MATCH($A2;$A$1:$A$5;0)
842020700000003	1	=MATCH($A3;$A$1:$A$5;0)
842020700000004	2	=MATCH($A4;$A$1:$A$5;0)
842020700000005	3	=MATCH($A5;$A$1:$A$5;0)

Expected results: 1;2;3;4;5
Comment 1 jbf.faure 2009-07-05 08:14:51 UTC 
Please be more explicit : 
- describe step by step what you do to encounter the problem.
- attach a sample document
Comment 2 villeroy 2009-07-05 11:08:35 UTC 
Created attachment 63366 [details]
Values and matches
Comment 3 Regina Henschel 2009-07-05 14:43:50 UTC 
The problem is not only in MATCH and VLOOKUP, but it is in normal comparison =
and operator - too, see attached file.

I personally think, that working with pure integrals OOo should be exact in
whole domain of 15 decimal digits. It is possible with IEEE double type.
Otherwise the domain for integrals has to be documented.

There has been a similar problem in issue 69798, which was closed as duplicate
of issue 69749. I think, that was a wrong decision, because there is no rounding
involved.
Comment 4 Regina Henschel 2009-07-05 14:46:19 UTC 
Created attachment 63367 [details]
wrong results for = and - with large integral values
Comment 5 Regina Henschel 2009-07-05 17:44:28 UTC 
Created attachment 63372 [details]
The difference is right, when calculated via macro, but wrong when calculated directly. The error is not in IEEE but in OOo.