DrawingML and SharedMLs have different text representations compared with word processing, spreadsheets and presentations definitions. This will cause wrong data formats and data loss if an application describes content based either first or second description.

Avoid the use of DrawingML and SharedMLs due they are not XML compliant specifications and have limited backward compatibility for text representations within proposed standard.

Clauses 5 and 7 Part 3

ge

Proposed Disposition of DIS 29500 Comment MX-0003 (Modified: 2008-01-10) DIS 29500 does indeed use similar text models for each of the three document types. The common architecture includes an outer paragraph element (or string instance, in the case of SpreadsheetML), nested run elements, run properties, and text nodes inside the runs. There is a need though for the three models to diverge in certain areas. The use of rich text within a word-processing document is quite different from that within a spreadsheet. Text within a presentation might have animations applied to it, while text within a word-processing document deals with functionality like page number fields or section breaks. Throughout the specification, we have allowed for future innovation and evolution within each of the three document types through the use of type-specific models that are not constrained by limitations or requirements of other document types.