Changing software applications, GEDCOM is poor minimal-common-subset tool

[pfarrell]

Over in the Wishlist, I posted a wish for improved Web Reports using Responsive Design.
See: http://www.reuniontalk.com/showthread.php?t=10627

Many folks say that rather than hoping Leister to release an improved version, I should just use Package X or Y that can read Reunion GEDCOM files and do A, B, or C that I want. I really don't want that. I really hope that Leister will someday agree with me and support Responsive Design.

I've been challenged on exactly what GEDCOM is missing. I'm not into that level of discussion, because I learned 15+ years ago that contrary to what a lot of non-technical folks think, GEDCOM was never intended as a universal genealogy and history recording transfer tool.

I haven't looked into the issue recently, but the GEDCOM format hasn't been updated since 1995, so I doubt anything has improved. See Wikipedia:

https://en.wikipedia.org/wiki/GEDCOM#Limitations

The fundamental issue is that GEDCOM was designed by the LDS to meet their theological and dogmatic needs. They say it does that. I am not in a position to argue with them. The LDS folks have said that they have no interest in changing or enhancing GEDCOM to meet other folk's needs, again a position that I have no argument with.

What I do have an argument with is that GEDCOM is anything other than a least-common-subset of what genealogists need to transfer.

When I started doing genealogy back in the 90s or whenever, I used a package called Generations which was a Windows port of Reunion by some PC development shop. By 2003 or so, they killed the product (Generations) so I started looking for an alternative.

What I found was that there are zillions of genealogy programs that can exchange a lowest-common-denominator/subset of data, but the transfer loses key information outside the lowest subset.

Plus there are three very different philosophies behind the databases.

1) are people oriented, so everything is about Mary Maguire
2) are family oriented, so everything is about Chris Smith and Mary Maquire and family
3) are event oriented, so everything is about Chris and Mary getting married April 2, 2011
and Leona Smith born May 14, 2012

while you can express any person, marriage or event in any package, its a lot easier to do what matches the philosophy and if the philosophies match, transferring data is a lot easier to automate.

The reason that I am using Reunion now is that I searched through looking for a Windows package to use after Generations died. I bought a bunch of them and suffered through the learning curve. None of them could capture all of the information that I had worked so hard to collect. I have a combination of data from a variety of sources, including work of my sainted Mother done 50+ years ago. I have people oriented data, family oriented data, event oriented data and random folklore and history.

When I had to buy a Mac for work on a project developing software "apps" for the iPad, I picked up a copy of Reunion. The wonderful folks at Leister converted 100% of my old data to Reunion, and I'm a happy Reunion user.

My personal interest is not in the "begats" that the bible talks about. I don't really care who was married to who, which kids they have, etc. What I care about are the stories. I want to know which of my relatives were horse thieves, crooked politicians, scoundrels, etc. I just have to record the begats to know who they were.

I have found that process of picking a new genealogy package to be extremely frustrating. All of the packages do the common stuff. Most do the common stuff with a fair amount of flair. I found out how well or not-well they did the uncommon stuff only after putting in a lot of time and effort on the individual package's learning curve.

There is clearly an industry-wide need for a stronger format for data interchange. As Leister says, this is really a database, and there are tools to convert from say Microsoft SqlServer to Oracle that don't make the user worry about losing links. One group tried to create GEDCOM-X, there are other "open standard". But as a wag said long ago, the great thing about open standards is that there are so many of them.

[riegelstamm]

Re: Changing software applications, GEDCOM is poor minimal-common-subset tool

Responsive Design and GEDCOM seem like two separate issues to me. The first pertains to how information is displayed on a device; the second pertains to how information is exchanged between different systems. Clearly we need an open source standard for exchanging genealogical data. There's no reason a Reunion user shouldn't be able to send their data to a RootsMagic user without loss of data. Other fields do this, there's no reason genealogy can't. There are projects to improve on GEDCOM, like Better GEDCOM, but's it's slow going.

But back to Responsive Design, have you looked into something like The Next Generation of Genealogy Sitebuilding, Family Tree PHP, HuMo-gen, or some other GEDCOM to PHP program? Since they build web pages, perhaps one of them supports responsive design.

[pfarrell]

Re: Changing software applications, GEDCOM is poor minimal-common-subset tool

Responsive Design and GEDCOM seem like two separate issues to me.

They are, which is why this is a new, separate topic.

[pfarrell]

Re: Changing software applications, GEDCOM is poor minimal-common-subset tool

There's no reason a Reunion user shouldn't be able to send their data to a RootsMagic user without loss of data. Other fields do this, there's no reason genealogy can't. There are projects to improve on GEDCOM, like Better GEDCOM, but's it's slow going.

The reason you can't, is that GEDCOM is inadequate. Slow is optimistic for BetterGedcom and others. They don't seem to have much life.

My guess is that the economics are against it. Ancestry.com seems to suck most of the oxygen out of the community as a whole. The software industry is rapidly moving away from expensive desktop packages and towards nearly free "apps" for mobile devices. I'm not convinced that this is good for complex apps, like photo or video editing and genealogy, but that is where the industry seems to be putting its money and engineers.

The Reunion help has a large section on common limitations of GEDCOM:

GEDCOM Limitations
Over the years, the GEDCOM specification has been the root of many problems with genealogical data exchange. There are two primary reasons for this:

1. The specification was written in such a way that it is impractical for genealogy software developers to support all its possibilities. The GEDCOM specification and the PAF (Personal Ancestral File) software are both products of the LDS church, yet the implementation of GEDCOM in PAF is quite limited. In effect, the LDS church wrote the GEDCOM rules so anything is fair game, but wrote their own software to support only a tiny subset of its possibilities.

In addition, the specification (particularly version 4.0) contains many elements that are easily interpreted in different ways by different developers. Further, between the official releases of GEDCOM 4.0 and GEDCOM 5.5, several "draft" versions were created, and some programs utilize elements from draft versions in their GEDCOM features.

2. A general perception exists that GEDCOM was designed for complete data transfer between software programs. It was not. In fact, according to the LDS church, the role of GEDCOM is limited to transferring data in "common fields." Generally speaking, the transfer of data between fields that are common to different programs is fairly reliable. What is unpredictable, and often problematic, is the transfer of data tied to specific features of a particular program or data residing in fields that are exclusive to one program or the other.

The result is that GEDCOM, as a standard file format, is much less reliable than other standard file formats (such as TIFF, HTML, or RTF) as a means of exchanging or translating data. Moving data between different software programs via GEDCOM may not provide a complete and accurate translation of your data, and using GEDCOM does not reconcile differing data structures employed by different developers.

Every genealogy software program uses different methods of information management, has different capabilities and limitations, and may interpret or implement GEDCOM differently. These factors all affect the fidelity of information translated between genealogy programs.

Here we explain how to use Reunion to import and export GEDCOM files; however, there simply isn't enough space to review all the potential pitfalls of GEDCOM data exchange. You may contact us if you have specific questions about Reunion's implementation of GEDCOM. If you have questions regarding the implementation of GEDCOM in another program, you should contact the developer of the other program.