Pages

Tuesday, November 12, 2013

What ever happened (is happening) to GEDCOM X?

No, GEDCOM X is not dead. It is not even sleeping. It is alive and well.  Quoting from the documentation:

GEDCOM X is a specification for an open data model and an open serialization format to support genealogical data communications. The FamilySearch Family Tree API is built on this specification.

The world has used GEDCOM as a standard to exchange family data for many years. The problem with GEDCOM is that even though it has become a de facto standard, there is data loss as files are passed between systems that have differing features and supported options. The problem grows even deeper as the world picks up pace in building even more family history applications in an online social environment. It is time for a major facelift to the process.

In addition to text-based lineage-linked data, GEDCOM X is capable of preserving rich media content in a new file format that will offer high-fidelity preservation of records. But more than that, we live in a connected world, and we need a common protocol to communicate family data between cloud systems and apps. This is what GEDCOM X is all about. It is currently being developed by FamilySearch in cooperation with many genealogy and technical experts.

Of course, there are other efforts to establish a standard for genealogical data exchange, but part of the trick in establishing a standard is to become involved in an actual data exchange problem. FamilySearch has just recently announced the formation of three different arrangements with the other three large genealogical database companies. How are they going to do what they announced? That is, how are they going to exchange the data between FamilySearch.org and the other three; Ancestry.com, D.C.Thompson Family History and MyHeritage.com? Perhaps, the whole problem can be and is being solved with GEDCOM X?

You might not have seen any announcements but the first stable milestone of the core GEDCOM X specification set has been released. The purpose of the milestone release is to stabilize the specification set to focus on documentation and developer tools. It is highly likely that FamilySearch has already implemented GEDCOM X in this process of data exchange out of necessity. In these circumstances, it is impractical to wait for a community wide consensus. A community wide standard does not become a standard merely because someone says that it is. A standard becomes a standard when it is used by a large enough and influential enough segment of the community. Think about it. The influential part is starting to happen with GEDCOM X whatever you might think of it.


2 comments:

  1. Re your question "how are they going to exchange the data between FamilySearch.org and the other three?" - a good question indeed. That question splits into two - how will they exchange the data from the sources (e.g. the images or transcriptions) and how will they exchange the indexes (that point to the images or transcriptions). Given that indexes can be unique to the source datasets, it will be interesting to see if there is any ability to map to a common model (GEDCOM-X) that started out life modelling people not sources.

    I am in several minds about your comment "In these circumstances, it is impractical to wait for a community wide consensus. ... A standard becomes a standard when it is used by a large enough and influential enough segment of the community". I have seen enough attempts to gain consensus dry up to sympathise. Allowing the eight-hundred-pound gorilla in the room to dictate things might help. But - what guarantee do we have that FS won't do with GEDCOM-X exactly what they did with the original GEDCOM? Which was to dump it (because the LDS is not a software house) while not releasing it to the community that might have done something with it....

    ReplyDelete
  2. The announcement of the latest GEDCOM X version was in June. They appear to have settled on the Java version. See Tamura Jones' comments:

    http://www.tamurajones.net/GEDCOMX1.0.0M1.xhtml

    ReplyDelete