||
|
 Sunday, March 16, 2008
Representation vs interpretation vs bare faced reverse engineering
The guys @ DIN (the German standards body for you non-skiiers) have done a bunch of work on a guide for translating between OOXML and ODF. They're seeing some of the same issues that we have come across in terms of ODF being 'lightly' specified.
"2.2 Representation vs. interpretation
Both standards OpenDocument and Office Open XML focus on specifying the syntax or representation. However, to be able to define a mapping between the two standards a knowledge about the underlying interpretation or semantic of the different XML token and attribute is important. For illustration consider the following example: Both OpenDocument and Office Open XML allow the definition of tab stops for a paragraph. In OpenDocument this is done by the <style:tab-stop> element. The position of a tab stop is defined by the style:position attribute. Its documentation reads: ―The style:position attribute specifies the position of a tab stop.‖ In Office Open XML the tab stop is defined by the <w:tab> element and the position is specified by the w:pos attribute: ―Specifies the position of the current custom tab stop with respect to the current page margins.‖. The problem here is that OpenDocument does not specify whether the tab-stop position is relative to the margin or relative to the paragraph indent. (Please note that OpenDocument differentiates between tab-stops relative to margin or paragraph indent in the table of contents - but is silent about general tab stops. Also note that e.g. OpenOffice.org Writer treats style:position to be relative to the paragraph indent.). These kinds of ―unspecified behavior‖ make a precise mapping definition hard."
It's like the whole "it's unspecified but OpenOffice does it loike xxx" thing all over again.
PoliTechLaw|Sunday, March 16, 2008 10:54:35 PM UTC||
|
|
|