[Libreoffice-bugs] [Bug 127476] Rounding behaviour of YEAR and SECOND function is inconsistent

bugzilla-daemon at bugs.documentfoundation.org bugzilla-daemon at bugs.documentfoundation.org
Tue Apr 21 15:09:37 UTC 2020


https://bugs.documentfoundation.org/show_bug.cgi?id=127476

Albrecht Müller <albrecht.mueller at astrail.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEEDINFO                    |UNCONFIRMED
     Ever confirmed|1                           |0

--- Comment #6 from Albrecht Müller <albrecht.mueller at astrail.de> ---
I uploaded two sample files: an .odt and an .xls version. No matter of if I
open the .odt file with the an old 6.0.4.2 Windows version of LibreOffice or a
6.3.5.2 Linux version: both show 0 seconds and year 1999. I used the .xls
version to open it with a Excel 97 version which is more than 20 years old.
Excel showed the same values as LibreOffice but when I reentered the formula
that calculates the year the value changed from 1999 to 2000. So I assume that
the initial values are just those stored in the .xls file and the value 2000
comes from Excel’s own calculation.

My explanation of this behaviour: More than 20 years ago Microsoft engineers
were aware of the fact that when rounding 59.9 seconds to 60 and then mapping
this to 0 seconds a carry should go to the minutes value and this carry may
ripple down to the years position. LibreOffice seems to use no carry in this
case. Consequence: The two programs disagree into which millenium the result
should go.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/libreoffice-bugs/attachments/20200421/a30d6dec/attachment.htm>


More information about the Libreoffice-bugs mailing list