https://bugs.freedesktop.org/show_bug.cgi?id=67026

Owen Genat <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Other: Calculation error    |Other: Limitations in Calc
                   |                            |precision need to be
                   |                            |documented

--- Comment #4 from Owen Genat <[email protected]> ---
(In reply to comment #3)
> Perhaps related to or same as Bug 37923

Almost certainly a duplicate. Summary edited for clarity.

(In reply to comment #2)
> I changed my mind and reopened this bug. The reason is that I think this is
> something very important that should be put into the documentation. Can
> someone check the documentation if this is included and add if it is not,
> add it. I will change the component from Calc to Documentation.

Agreed. Can someone add the Documentation mailing list to the CC list for
greater visibility and to solicit comment? I tried and failed. I had a look
through the Calc Guide v4.1 and note:

- "precision" references are mainly to the "Precision as shown" option (pp.222,
361-362) and the EUROCONVERT() function (p.388). I note the entry for
"precision" in the index indicates p.354, which seems unrelated. Also, note
that the "Precision as shown" option appears dependant on the cell format not
being General (it must be a Number) as recently discussed here (in the comment
by Mark Bourne):

http://nabble.documentfoundation.org/Error-or-what-td4095174.html

- " rounding" (leading space to avoid some false positives) references are
mainly to the topic of "Rounding off numbers" (pp.221-222) and the "Decimal
places" display option when formatting cells (pp.362), as well as the FLOOR()
and CEILING() functions.

It would be good if this limitation, in how integer values are handled
(generally by all computers) affects precision in calculation, could be added
to the beginning of the Calc Guide somewhere (the Preface?). IMO this needs to
be clearly highlighted using the "Caution!" block style (shown on p.174) and
preferably present a clear example of basic addition and subtraction using
large and small values. This would enable the developers, QA, and others to
point out a location where this limitation is officially described.

I can't find any information related to this topic (using key terms such as
"Calc" + "precision", "rounding", or "limitation") in either the LO Help
website or TDF wiki. It should be added to these repositories also.

-- 
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Libreoffice-bugs mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs

Reply via email to