you are not logged in


User login

M-XML (Mix-em-mel)

M-XML is the dataformat defined for the MIXED software & plugins.

Currently, the part concerning spreadsheets is in progress.

The discussion about M-XML between Kris and Dirk has shown deep fundamental differences in the view upon this standard. Dirk feels that M-XML for spreadsheets should be as close as possible towards ODF, profiting from the developments done in this field, and enabling a maximal use of existing software for conversions. Kris feels that M-XML should be self-describing and as explicit as necessairy to enable a long-term preservation where we do not link our standard directly to a model which is not developed for long-term preservation. Kris also feels that the project intitiation document corresponds with his view.

At the end, Dirk, being projectleader for MIXED, has taken the decision and responsability to direct Kris into redoing the work done for M-XML with specific guidelines.

These guidelines are:

  • define the 'umbrella' for coexistence of different data-structures supported by MIXED
  • define the M-XML spreadsheet implementation of ODF, based on the ODF specification, and specify when our implementation differs from ODF

Dirk has also set specific guidelines for the database-model, being (formulation by Dirk):

  • the data-model (tables, fields, keys, constraints) must be modelled in M-XML
  • the data itself must be modelled in M-XML
    • first consideration: model the records and fields of the tables so that it has a natural fit with relational table data
    • second consideration: try to get as close to M-XML for spreadsheets as possible, but the first consideration takes precedence
  • keep the M-XML fragment for the datamodel separate from the m-XML fragment of the data

Current version: Standard Data Formats for preservation document (v0.3)

Standard Data Formats for preservation document (v0.2)

SDFP document (v02b).

SDFP schema (v04).


M-XML 6-8-07.doc93.5 KB
M-XML current.doc202 KB

ODF not designed for long-term perservation

Although long term preservation is not the primary goal of ODF, it turns up more and more in discussions. It is mentioned in Microsoft's motivation for Open XML, and it is put into practice already.

I see the ODF and OpenXML standards as something that potentially achieve the goals of MIXED, but then for Office documents in general. Even if they are not yet build for preservation, the need for that will be felt, and something will be done about that.

Concerning the properties self-describing and explicitness: I do not think the documentation should be in the tags, but in the XML-schema. In the end, it will be computers that read our archived documents. The only humans that need to interpret M-XML documents are those who must develop new algorithms to access their content. We may assume that if they have M-XML documents, that there is also an M-XML schema around to help them.