DR 16-0007

Thu Dec 14 01:40:04 CET 2017

Dear all

I have an action from today’s WG 4 telecon to confirm the prose changes that
I believe need to be made in order to close DR 16-0007. All schema changes
that I proposed have already been made by Murata-san, or so I believe.
Caroline has already indicated that she is happy with some of my proposed
prose changes. There are some outstanding questions to the Excel team for
which we still await responses, so this is still work in progress, but here
is the current status regarding prose changes.

§18.2.2 calcPr (@forceFullCalc)

On reflection, if we add the default value “false” to the schema, I don’t
think we need to change the prose. LibreOffice does not support this
attribute, i.e. it ignores it on consumption and throws it away when the
document is saved, so I think that the behaviour of Excel is definitive.

§18.2.7 ext (@uri)

The proposal was either to add the words “If present,” at the start of the
description of @uri, or take no action. On August 25th 2017 Caroline said
“The proposed resolution looks good to me. Assuming we make any changes as a
result of this DR, I don't see any problems with making the changes Francis
suggests "for clarity."”. That seems to steer us in the direction of adding
the words “If present,”.

§18.3.1.73 row (@r)

I proposed the following be added to the description of @r:

If this attribute is not specified, the row index shall be interpreted to be
1 greater than that of the previous row in the parent element. If this
attribute is not specified and the row is the first in the parent element,
the row index shall be interpreted to be ‘1’. 

This should go at the end of the first paragraph of the description. No-one
has commented on this proposal, so I suggest we agree to make this change.

As a consequence, for consistency, I propose the following change to the
description of @r on §18.3.1.4 c (in accordance with the corresponding
implementer note in [MS-OI29500].pdf):

If this attribute is not specified, the column index part of the cell
reference shall be interpreted to be one greater than that of the previous
cell in the parent row element. If this attribute is not specified and the
cell is the first in the parent row element, the column index part of the
cell reference shall be interpreted to be ‘A’.

§18.3.2.10 top10 (@filterVal)

No change to normative prose has been suggested, but we should correct the
Example as indicated in the DR log.

§18.8.7 cellStyle (@name)

My original proposal was to add the following text to the end of the first
sentence of the attribute description:

If this attribute and builtinId are both omitted, this cell style shall be
ignored.

However, following further tests in Excel and LibreOffice, I conclude that
the behaviour is implementation-dependent. For example, <cellStyle
xfId=”0”/> is acceptable to both Excel and LibreOffice and in both cases is
saved as <cellStyle xfId=”0” name=”Normal” builtinId=”0”/> (assuming that
there is not already a style of this name. However, <cellStyle xfId=”1”/> is
acceptable to both Excel and LibreOffice, but Excel saves this as <cellStyle
xfId=”1” name=”Style 1”/>, while LibreOffice discards the style. I therefore
propose instead:

If this attribute and builtinId are both omitted, the behaviour is
application-dependent.

§18.10.1.2 b (@cp, @f, @u)

I did not propose any change to the prose, except to suggest that an example
of the use of @cp on one of the elements b, d, e, m, n or s would be
helpful. If no-one in this group has sufficient expertise with Pivot Tables
to provide an example, we may need to refer this to the Excel team. The
suggested question to the Excel team would be: 

Can you provide one or more suitable examples of the use of the attribute cp
on any of the elements b, d, e, m, n or s, to indicate the number of
properties of the item, for inclusion in the attribute description table of
one or all of these elements in §18.10.1.2, §18.10.1.21, §18.10.1.27,
§18.10.1.50, §18.10.1.60 or §18.10.1.85.

§18.10.1.8 calculatedItem (@formula)

We’re awaiting a response to my question to the Excel team.

§18.10.1.21, §18.10.1.27, §18.10.1.50, §18.10.1.60, §18.10.1.85

See §18.10.1.2 above.

§18.10.1.45 item (@x)

We don’t have a resolution to this. I’m wondering whether we should move all
the preceding Pivot Table related elements/attributes from this DR into a
new DR, so that we can reach a resolution of the remainder more quickly. 

§18.10.1.86 serverFormat (@culture, @format)

My proposal was to add the following text at the end of the paragraph above
the attribute description table:

At least one of the attributes culture and format shall be specified. If
both attributes are specified, the result is application-dependent.

I have nothing further to add, so this remains my proposed resolution.

§18.10.1.92 tpl (@fld, @hier)

I suggest a new question to the Excel team in order to try to resolve this:

What is meant by “Office requires that @fld and @hier be mutually exclusive
in SpreadsheetML documents”? Does this mean that a document may only contain
one or other of these attributes? Or does it mean that each tuple may only
contain one or other of these attributes, but some tuples may contain one
while other tuples contain the other attribute?

§18.13.3 dbPr (@command)

We’re awaiting a response to my question to the Excel team.

§18.13.12 textPr (@characterSet)

My proposal was to change the description of @characterSet to read:

If this attribute is not present, the behaviour is application-dependent.

There has been no objection to this proposal, so I see no cause to revise
it.

§18.13.13 webPr (@url)

For clarity, I propose adding the text I proposed previously to the
description of @url:

This attribute is required, unless the attribute deleted of the parent
connection element is specified to be true.

§18.14.5 definedName (@refersTo)

We’re awaiting a response to my question to the Excel team.

That covers all the elements for which we have discussed possible prose
changes.

Regards,

Francis

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.vse.cz/pipermail/sc34wg4/attachments/20171214/d53d4082/attachment.html>