[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-16 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

Heiko Tietze  changed:

   What|Removed |Added

   Keywords|needsUXEval |
 CC|libreoffice-ux-advise@lists |
   |.freedesktop.org|

--- Comment #33 from Heiko Tietze  ---
(In reply to V Stuart Foote from comment #30)
> We may annoy a few users, but there needs to be clear UX agreement of the
> approach to correcting this. To me a clean break is better than muddled
> attempt to accommodate legacy Writer table cell addressing.

ODF conformity is prime likewise compatibility. It should be possible to open
documents with old/wrong and new/correct references in upcoming and old
versions.

OTOH, whether a1 is the first or the 27th column would be unclear to users. I
agree that most tables are shorter. So perhaps a clear cut is better than
fiddling around with compatibility.

But that's rather a decision for the ESC. And we gave enough input.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-12 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

--- Comment #32 from Eike Rathke  ---
(In reply to Dave McKellar from comment #31)
> Changing Writer cell addressing to be the same as Calc means the code could
> be shared too.
Wishful thinking, but no, it does not. They are completely different
implementations and suit different needs.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-11 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

--- Comment #31 from Dave McKellar  ---
Changing Writer cell addressing to be the same as Calc means the code could be
shared too.  Besides slightly reducing the codebase this has a UX benefit: the
two applications will be in sync for future improvements or bug fixes.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-10 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

V Stuart Foote  changed:

   What|Removed |Added

   Keywords||needsUXEval

--- Comment #30 from V Stuart Foote  ---
(In reply to Heiko Tietze from comment #29)
> Issue seems to be clear (with a lot of facets). If we keep lowercase and
> accept =sum() it might break backward compatibility.
> 
> Removing needsUXEval.

Sigh, "fixing" this would require major refactoring of Writer tables. Where
using =sum() is perfectly acceptable cell reference to 1st, 2nd and 3rd
row of the *27th* column of a Writer table! If they don't exist in the table,
it alerts.

In a Calc sheet those cells would be referenced by . And, entering
that same range in lower case is corrected by Calc edit shell to . ODF
standard does not permit addressing cells with lower case--they are converted
in Calc.

Calc complies with ODF, Writer does not.

But this is an UX issue requiring assessment of what extent do we need to
accommodate the prior usage?  And how quickly this needs to be corrected? 

UX Considerations:
1) how common are tables in our Writer .ODT wider than 26 columns?
2) how disruptive would it be changing Writer formula input to ODF standard?
3) can we let the current addressing remain unchanged?
4) are we obliged to maintain backward compatibility of Writer table addressing
for use on prior builds?
5) is it better to fully convert a Writer document to new "corrected" ODF
compliance, and when (on opening, or on save)? Or should we attempt more
complex handling of supporting both addressing schemes?


To me it is pretty simple. "Zero" complaints from OOo/AOO/LO means few if any
users of Writer have bumped into the issue of mixed-case addressing in Writer
tables wider than 26 columns. They are not common!  Not much disruption by
fixing it correctly to comply with the ODF document model.  Frankly we must
correct it (or the world questions our standards compliance--we have enough
issues with SVG already).

The last two are a bit tougher, and interlinked. I'd argue strongly that the
best action is to aggressively refactor handling of formula addressing of
Writer tables to be ODF table cell address compliant--eliminate lowercase 52*2
addressing. Then match the Calc shell validation to convert lowercase -> upper
case cell ref input.  

For simplicity .ODT filter import should convert incoming Table cell formulas
from the A-Z,a-z,AA-AZ cell addressing to ODF compliant A-Z,AA-AZ,BA-BZ. And we
should write back any formulas in ODF compliant form. The document save will
complete the conversion--only closing without save should abandon conversion.

Any new Writer table formulas will be correct, any old documents would be
irrevocably converted. Yes someone on an old build of LO opening a new document
would have formula issues--but then the old version is not an ODF compliant
renderer.

We may annoy a few users, but there needs to be clear UX agreement of the
approach to correcting this. To me a clean break is better than muddled attempt
to accommodate legacy Writer table cell addressing.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-10 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

Heiko Tietze  changed:

   What|Removed |Added

   Keywords|needsUXEval |
 CC||heiko.tietze@documentfounda
   ||tion.org,
   ||vmik...@collabora.com
   Severity|enhancement |normal
 Blocks|120200  |122755

--- Comment #29 from Heiko Tietze  ---
Issue seems to be clear (with a lot of facets). If we keep lowercase and accept
=sum() it might break backward compatibility.

Removing needsUXEval.


Referenced Bugs:

https://bugs.documentfoundation.org/show_bug.cgi?id=120200
[Bug 120200] [META] Update or correct help for features
https://bugs.documentfoundation.org/show_bug.cgi?id=122755
[Bug 122755] [META] Formulae in Writer tables bugs and enhancements
-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-05 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

--- Comment #28 from Mike Kaganski  ---
(In reply to Eike Rathke from comment #27)
> It's still a bug because ODF 9.2.1 Referencing Table Cells
> https://docs.oasis-open.org/office/OpenDocument/v1.3/os/part3-schema/
> OpenDocument-v1.3-os-part3-schema.html#__RefHeading__1415614_253892949
> says that after Z the next column 27 is AA.

Yay! :-D

> Good luck ;-)

I second this.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-05 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

--- Comment #27 from Eike Rathke  ---
(In reply to Mike Kaganski from comment #17)
> Heh, so it is NOTABUG.
> Try to insert a table with e.g. 60 columns, and check its columns :-)

It's still a bug because ODF 9.2.1 Referencing Table Cells
https://docs.oasis-open.org/office/OpenDocument/v1.3/os/part3-schema/OpenDocument-v1.3-os-part3-schema.html#__RefHeading__1415614_253892949
says that after Z the next column 27 is AA.

However, if this is corrected then
a) import must take care of converting everything above A-Z to the proper
notation
b) writing the proper values of AA... to file will let older versions address
wrong columns
c) when reading file and encountering AA... it needs to decide whether that
could be the old broken notation for value 53... or the new 27..., likely by
inspecting whether it fits into the actual table width
d) though not recommended and frowned upon, it may be necessary to act on the
 value if that is neither LibreOffice nor OpenOffice then
assume the proper ODF conform values were written; if it is LibreOffice then
its version number could help to determine how to handle c)

So, to remedy b) somewhat, when fixing this in master apply changes for current
release branches that when encountering AA... do c) but do *not* the conversion
of a) and b) to not propagate the fixed values to even older release versions
that don't know anything about this.

Good luck ;-)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Libreoffice-ux-advise] [Bug 145384] Writer table cells have address scheme that differs from Calc, "A-Z,a-z,AB-AZ,..." vs "A-Z,AA-AZ,BA-BZ,..."; lowercase cell refs are more error prone for cell form

2021-11-04 Thread bugzilla-daemon
https://bugs.documentfoundation.org/show_bug.cgi?id=145384

V Stuart Foote  changed:

   What|Removed |Added

   Keywords||needsUXEval
Summary|Lowercase cell references   |Writer table cells have
   |not tolerated in formulas   |address scheme that differs
   |for Writer table cells  |from Calc,
   ||"A-Z,a-z,AB-AZ,..." vs
   ||"A-Z,AA-AZ,BA-BZ,...";
   ||lowercase cell refs are
   ||more error prone for cell
   ||formulas
 CC||libreoffice-ux-advise@lists
   ||.freedesktop.org

-- 
You are receiving this mail because:
You are on the CC list for the bug.