https://bugs.documentfoundation.org/show_bug.cgi?id=107850
dumb...@gmail.com changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|CLOSED |RESOLVED
--- Comment #6 from dumb...@gmail.com ---
Hi Eike,
thank you for your time and an answer shedding some light on this situation.
(In reply to Eike Rathke from comment #5)
> (In reply to dumblob from comment #0)
> > 7. Change content of cell A1 to: "Sheet1X.A5 + Sheet1.$A$6" and press Enter.
> > #REF! appears instead of the number "5".
> It has to because a sheet named Sheet1X does not exist.
Of course, it's correct (I didn't write it's a bug ;) ).
> > 8. Click on the cell A1 and look at the content - "Sheet1X.A5" changed to
> > "#REF!.A5". This is the first bug (the correct behavior would be to leave
> > the cell content as it was).
> That's debatable. If the reference string was left as it was entered
> wrongly, in a longer formula expression you wouldn't even see which part of
> the expression was wrong.
It might be debatable, but the fact is, that it destroys a lot of work and
therefore I can't see any other label than a "serious usability bug". There are
plenty ways to show which part of the formula is causing issues (font
formatting like underline or strikethrough or a different color or a different
font etc. or my favourite: just adding (not replacing) the "#REF!" string at
the end of the problematic reference in the formula...). See my questions
below.
> > 9. Save the document (again to "test00.ods").
> > 10. Close all running LO instances and open Calc and then the file
> > "test00.ods".
> > 11. Click on the cell A1 and look at the content - "#REF!.A5" changed to
> > "#REF!". This is the second bug (the correct behavior would be to leave the
> > cell content as it was).
> It has to be #REF! because the ODF OpenFormula specification specifies that
> an invalid reference is to be written as #REF!. Earlier versions wrote the
> more detailed but wrong #REF!.A5 string, that has been fixed which seems to
> be viewed as regression. It is not.
Let me express my opinion. Let me also apologize in advance for my strong words
(I'm aware you're one of the ODF authors).
This behavior seems utterly wrong and absolutely misleading. There seems to be
no sane reason to throw away such an important part of a formula (a particular
reference having been invalidated by actions that occurred after the formula
was validly created). I strongly believe, that such a nonsense is not defined
in the ODF specification.
Could you please point me to the relevant parts of the ODF specification? I
found "#REF!" in
http://docs.oasis-open.org/office/v1.2/OpenDocument-v1.2-part2.html
with a description not clearly distinguishing between the formula itself and
the visible result shown in the cell. Then I found
http://docs.oasis-open.org/office/v1.2/os/OpenDocument-v1.2-os-part2.html#__RefHeading__1017900_715980110
with the same impression. I could not find definitions of "constant Error",
"Error value", "original specific Error name", "Error constant", "Error", "data
value #N/A is an error value", "error information". These terms are
unfortunately used interchangeably in a quite messy manner.
According to my understanding, the OpenFormula spec does not dictate any
replacing of the erroneous reference with the string representing the error
which occurred.
More importantly, could you also point me to the correct place where I could
raise this issue with suggestions? Some errata bug tracker or similar for the
ODF specification.
--
You are receiving this mail because:
You are the assignee for the bug._______________________________________________
Libreoffice-bugs mailing list
Libreoffice-bugs@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs
