RE: Office 2003 Beta

[Ann Hastings <Ann@xxxxxxx>]

Yes, this seems to be the problem, as one responder succinctly put it:

Re:You're Wrong As Is The Article (Score:5, Interesting) by Tailhook (98486)
on Thursday March 13, @02:15PM (#5505335) 

"Basically it looks like the authors of the article want to have their cake
and eat it too. They somehow want to preserve all the formatting information
in their documents in the XML output yet not end up with a lot of Office
specific content in their documents."

The choices then appear to be "data only XML" or "RTF marked up XML". Is
this correct?

If so, then I think the critics are correct. The critics wish that the
document can be read and manipulated by some non-Microsoft editor. I doubt
this is feasible with the WordML format, aka "RTF marked up XML". I'll
explain why.

If, as you point out, a WordML document is collection of data marked up by
XML tags that provide only low-level (RDF) presentation information, then it
is of no use to an alternate editor implementation. An analogy would be to
attempt to edit a Word .doc document after saving it as RTF. The appearance
of the RTF when rendered is correct, but the data model that Word uses
internally is not represented in the RTF document. This is a "one way"
conversion. You can edit the RTF, but you can not reproduce the .doc file
from it.

It sounds to me like it is the same with WordML. You can read and edit
WordML because it's valid XML. However, the higher level data model of Word
is simply lost. No means is provided for a processor to understand the
original structure of the document.

For example, if I should create a "style" in Word and apply that style to a
paragraph, the WordML output will tell me what font to use. However, the
WordML file tells me nothing about the "style". So I can't tell what other
paragraphs are supposed to change in sync if I change the style. I can't
know the inheritance of style parameters. In short, I can't programmatically
edit the Word data model.

The hope/expectation was that the XML output would provide this information.
Thus, it would be possible to essentially re-implement the Word data model
and correctly manipulate Word documents. With this hope/expectation in mind,
it's clear why what they have actually found is considered crippled. WordML
can't be used to recreate any part of the original data model in an
alternate editor. It's just data mixed up with low level markup.

I have always thought that this expectation is niave. Microsoft protects the
tools they sell by making it infeasible to create alternate implementations.
Just because the tool can output XML doesn't mean that you can do without
Word.

BTW, I am by no means an expert at any of this. I'm no Office beta tester
and I haven't looked at OpenOffice in months.


-----Original Message-----
From: DW Emory [mailto:danemory@globalcrossing.net]
Sent: Friday, March 21, 2003 1:57 PM
To: Mark Barratt
Cc: Free Framers; FrameSGML List
Subject: Re: Office 2003 Beta

I guess my conclusion was that Microsoft is producing some sort of XSL that 
does not conform to the XSL standard, which is Microsoft's typical 
"solution," thereby forcing everyone to continue using Microsoft products. 
Perhaps I'd better investigate further.

** To unsubscribe, send a message to majordomo@omsys.com **
** with "unsubscribe framers" (no quotes) in the body.   **