[odf-discuss] What is actually necessary and found only in OpenXML?

Daniel daniel.carrera at zmsl.com
Mon May 28 08:11:54 EDT 2007


Marbux says:
>> So as to ODF 1.0, StarOffice and OpenOffice produce non-conformant ODF
>> because they destroy the foreign elements and attributes. But because 
>> of the
>> change in ODF 1.1, there are virtually no conformance requirements in the
>> entire specification. So Sun's ODF is now conformant even though its apps
>> destroy the foreign elements and attributes.

One second Marbux, stop here. I haven't been following this thread, but 
what the heck are you talking about? Is this about section 1.5 again? I 
have the OpenDocument spec version 1.0 in front of me. Tell me:

1. Exactly where does it say that applications are required to preserve 
foreign elements? Be concise and specific.

2. Exactly what changed between 1.0 and 1.1? Be concise and specific. I 
have versions 1.0 and 1.1 of the spec in front of me and on the topic of 
conformance the only difference I see is the change from the words 
"MUST" to "SHALL" which is preferred by ISO. Other than this, section 
1.5 "Document Processing and Conformance" is word-for-word identical.


In addition, I'd like to point out that preserving foreign elements is 
not, in general, a reasonable requirement. Suppose you have the following:

<text:p foreign:foo="432">Paragraph Alice</text:p>
<text:p foreign:foo="437">Paragraph Bob</text:p>

Then a user inserts a paragraph in between those two. What should the 
result look like? Do you copy the foreign element to the new paragraph? 
Not copying it might effectively break it (e.g. if this element denotes 
a list or some other grouping) but copying it might break it (and what 
value would you pick?).

There are special cases where preserving foreign elements might be a 
reasonable requirement, but this is not true in the general case.

Lars says:
> Just to be clear, does that mean that StarOffice and OpenOffice.org do 
> not yet create ODF-compliant documents?  If so, could you please point 
> to specific details?

I second Lars' request for specifity. I'm not aware of any instance 
where OOo produces an invalid document. I'm aware of instances where I 
disagree with how OOo displays a document, but that is different.

Marbux says:
>> That is a significant barrier to interoperability ...

That's like saying that OOo's inability to read .doc files perfectly is 
a significant barrier to interoperability.

Daniel.



More information about the odf-discuss mailing list