https://bugs.documentfoundation.org/show_bug.cgi?id=136615
--- Comment #40 from Regina Henschel <rb.hensc...@t-online.de> --- (In reply to Eike Rathke from comment #39) > (In reply to Regina Henschel from comment #38) > > That's horrible. It guarantees that if the parameter was used, > interoperability will fail with *ALL* implementations that don't implement > it, i.e. *ALL* current implementations. > Furthermore, for SECOND() with [ ; Logical Truncate = TRUE ] importing > existing documents (Excel or not, any) without that argument change the > behaviour, and exporting without will change behaviour in implementations > not knowing the parameter, i.e. probably all currently existing > implementations (because they implemented the sick Excel behaviour mandated > by the current ODFF specification). I do not see that. If the parameter is missing, the functions behave as specified in ODF 1.3. Only the contradiction in ODF 1.3 is removed, that ODF 1.3 on one hand specifies to round in the SECOND function resulting in range 0 to 60, and on the other hand specifies a range of 0 to 59 for the SECOND function. The other functions all truncate in ODF 1.3. So applications having implemented ODF 1.3 should produce the same results as in ODF 1.3, when they get an ODF 1.4 document without parameter. I think, that is a good backward compatibility. But that (changing behaviour) is the > same situation as it would be with the following simple redefinition: > > I suggest to rather only redefine SECOND() to always truncate and keep the > already existing (truncating) definition for all other functions touched in > that proposal, so values are never unexpectedly rounded into the next higher > unit, worst case the year after. How would you then solve the request to have functions, that behave like Excel? -- You are receiving this mail because: You are the assignee for the bug.
