OpenGen Data Model
[These are just rough notes to get an entry started for this model -- Tom Wetmore.]
OpenGen is another effort to define a standard for genealogical interchange. They have a website site at opengen.org.
I attended the OpenGen meeting held on 8 February 2011. The main topic was a review of a spreadsheet comparing the contents of GEDCOM and the contents of the developing OpenGen model.
The model will be published at RootsTech later this week and weekend. If you are a member of the OpenGen forum you can view both the comparison document and a current diagram of the model. Here is a diagram of the model as it existed on the OpenGen web site on 8 February 2011:
The model is very similar to other models that have been reviewed here in the Better GEDCOM wiki.
Specifically, the model includes Persons, Events, Sources, Places and Notes as first class citizens (that is, as record level objects). There are other components, but these, in my opinion, are the key components, and OpenGen contains them all.
As seems de rigueure
these days, the OpenGen model does not include a Family record, nor does it seem to include a Group record. Families must be reconstructed by software using information about parents and birth order that are maintained in the Person records.
I got the feeling from the meeting that the Evidence and Conclusion model is not on the OpenGen horizon, though I could be wrong. In other words there is no distinction between Persons that come directly from evidence, and person that are built up by a genealogist by making inferences from the the persons directly derived form the evidence. In this sense it seems that the OpenGen model is almost identical to the Gramps model in both structure and in intent. That is, Person records are built up by having them refer to more and more Event records. Though I'm not entirely positive about that, since nothing at this level seems written down about OpenGen at this point.
A few other points. The external file format is to be XML. The character set is not specified other than being any character set supported by the XML standard. Right now they are going with the "one way to do it" dictum that Better GEDCOM has also expressed, meaning for practical purposes that Sources, Places, and Notes must always be in separate records and cannot be tucked up into other records when that would be more appropriate. I disagree but it's not a federal case.
The point for Better GEDCOM to take from this is that the OpenGen organization, a definite "competitor" to Better GEDCOM in some sense, since overall goals are the same, to become the new standard for archiving and transporting genealogical data, is already deep into model building, whereas work on the Better GEDCOM data model is not yet officially underway.