Modern Software Experience


Yet another genealogy standard?


Last Thursday through Saturday, FamilySearch organised the RootsTech conference. On Saturday, Renee Zamora tweeted BREAKING #RootsTech news - FamilySearch's SWORD project is working on the next generation GEDCOM. Community assisted lead. and her tweet was quickly answered by requests for more information, but there was no more information to be had.


I've been in contact with OpenGen and BetterGEDCOM people, and during the  BetterGEDCOM Developer's Meeting today, Sandy Rumble summarised the meeting she attended. This articles combines what I've learned from various sources. It is not unlikely that some details are wrong.


The announcement was made twice. It was initially made on Thursday, during an unannounced break-out session. Few people knew about it, so few people attended, and those that did, did not blog or tweet about it.
On Saturday, a larger group was present for the scheduled Genealogical Data Standard discussion panel. This was a two-part session, and many who attended the first part did not return after the break. It was then that Tom Creighton, the FamilySearch Chief Technology Officer, made his announcement. After that, the panel discussion devolved into a press conference with Tom fielding questions from those present - and largely telling them that he could not answer those questions. Still, a few things are known.

soft announcement

First of all, Tom did not make a formal or official announcement, but what he called a soft announcement. It's a pre-announcement telling us that they are likely to make a formal announcement at a later date. The official announcement will follow when there is internal agreement, and the internal discussion is apparently about intellectual property rights.


Tom did not pre-announce SWORD, but SORD. This is said to be an acronym for SOurce Record Data. This is the name of either a data model, a file format, or both. There was talk about sharing some source code as well. There are no details about the nature of that code, but is likely to consist of some examples or reference code.

GEDCOM replacement

For years now, FamilySearch has been telling people that GEDCOM is dead, and that the NFS API is the alternative. There is a lot wrong with that claim. Even if NFS wasn't still in Alpha, but out of Beta already, it still wouldn't be true. Genealogists want to exchange data directly with each other without having to go through an intermediary, and are particularly unenthusiastic about an intermediary that intends to hang on to all their data.

Last year, both OpenGen and BetterGEDCOM were founded, two organisations that both seek to create a replacement for the GEDCOM standard that FamilySearch abandoned. In fact, BetterGEDCOM was formed in response to FamilySearch reiterating once again their position that they abandoned GEDCOM yeas ago and that they consider the NFS API to be all the GEDCOM replacement we need.

Now, as these two organisations are gaining traction, as a community is forming around these initiatives, and they are starting to make some progress, FamilySearch suddenly announces that they have a new data model and file format they'd like everyone to adopt? That's certainly remarkable.


During Saturday's impromptu press conference, Tom made it clear that NFS is a FamilySearch product, and SORD is a FamilySearch product, but that SORD is not from the NFS people. The group working on SORD has some philosophical differences with the NFS group. The most obvious difference is that the NFS people wants to dispense with file formats for exchange of genealogical data in favour of the NFS API, while the SORD group wants to have a data format.
The two groups are so independent of each other, that there is not even a one-to-one correspondence between their data models.


The SORD model is an object model, said to be influenced by the GenTech model. That is more or less what FamilySearch said when it introduced GEDCOM XML back in 2002, and that makes we wonder whether SORD isn't simply the current name of what their GEDCOM XML thinking has become.

industry standard

Although the SORD group naturally would like everyone to adopt their standard, FamilySearch is apparently not seeking to own the SORD standard like they own the GEDCOM standard, but seeking to have it managed by an industry standards body. FamilySearch does intend to participate actively in such body, but would no longer be the owner of the specification. That does seem the right approach; it would be certainly easier for an industry standards body than any single organisation to get their specification approved as an ISO specification.

standards organisation

We already have two standards organisations; OpenGen and BetterGEDCOM. It is not clear whether FamilySearch is seeking to establish a third organisation, or to submit their standard for consideration to either of the two existing ones.

It is not clear how the existing organisation will respond to the SORD proposal either. Both were surprised by the SORD announcement.
During the BetterGEDCOM meeting, I suggested to invite Tom Creighton to join in a BetterGEDCOM meeting, and it was decided to do so.



The preliminary RootsTech 2012 schedule includes the FamilySearch talks New gedcom (how to produce and consume it) and New gedcom (what it is, what's it's scope, how is the project managed and maintained?). Apparently, FamilySearch is planning to introduce New GEDCOM at RootsTech 2012.