GENMSC-L Archives

Archiver > GENMSC > 2007-07 > 1184720417


From: Wes Groleau <groleau+>
Subject: Re: Beyond GEDCOM
Date: Wed, 18 Jul 2007 01:00:17 GMT
References: <slrnf9ol6e.cvf.usenet@goodwill.larseighner.com>
In-Reply-To: <slrnf9ol6e.cvf.usenet@goodwill.larseighner.com>


Lars Eighner wrote:
> Here is the problem: GEDCOM has become the de facto standard for family
> history and geneology data. And the problem with that is that the Mormons
> own GEDCOM and thus it is rife with built-in religious limitations.

I too have lamented the notion that there are only two kinds of
relationships: marriage and child of a marriage. Finally, the
LDS realized that was a flaw and introduced "associations"

And then there's the 6.0 XML version--while introducing a completely
new lexical/syntactic structure, they had the opportunity to correct
the semantic inddequacies of the previous model. Instead, they
carefully ensured there was no practical difference and that the
conversion old format to new and vice versa is almost as trivial
as print to Morse code.

Is there any software out there that used the "event-linked"
GEDCOM model? I'd like to see a "relationship-linked" model:
Keep everything currently in GEDCOM 5.5.1 but add a series
of tags representing numerous relationships.

0 @person1@ INDI
....
1 FATH @person2@
1 MOTH @person3@
1 SIBL @person4@
1 WIFE @person5@
2 MARR @event1@

And why this devout slavery to the four-character abbreviation?
They already have FAM, AFN, CHR, & SEX so they know there's
no magic in the number four. In PLAC, BIRT, DEAT, what does
saving a single letter buy us other than making things just
a smidgen more difficult for people who aren't English?

> So, I would be grateful for any pointers to existing software that handles
> all families or suggestions for extensions.

A "FAM" should be a list of people who lived together,
with their roles, the time period it existed, and the
events that are significant to it:

0 FAM @family1@
1 HUSB @person2@ {{ John /Doe/ }}
1 FATH @person2@
2 NOTE FATH & HUSB could be interchangeable,
3 CONT or they could be different people--
3 CONC and there could be more than one
3 CONT (hippie commune of the sixties?)
1 MARR
2 DATE .....
2 HUSB @person2@
2 WIFE @person3@
1 MARR
2 DATE ....
2 HUSB @person2@
2 WIFE @person9@
2 NOTE Why not? Polygamy has existed. Plus,
3 CONT this would make it much easier to link
3 CONT a man to two wives without making half-
3 CONC brothers who lived together most of their
3 CONT lives into strangers as far as the file
3 CONT can show. Finally, when you find that
3 CONT last scrap that makes these two women
3 CONT the same person, how much easier is it
3 CONT to merge them within this FAM instead
3 CONT of merging two level zero FAMs ?

No doubt other people have put a lot more thought
into it than I have, and can find plenty of flaws
in my idea. But you asked for ideas, and this is
one that has been on my back burner for years.

Also, a "comment" syntax would be nice--my {{ John /Doe }}
above. Comments are like in programming, in a way--they
are information to the human reader that the software does
not use. However, they _could_ be more. Software could
generate them, and/or use them for an integrity check
if they are done with some consistent style.

I have some perl code that loads a GEDCOM into a Berkeley DB,
easily brings one level zero record out at a time unchanged
from how it went in. During output, it scans for Xrefs,
retrieves information on the record pointed to, and creates
it as a comment in the form above. This helps me as I edit
the GEDCOM. After editing, the software strips off the comments
before replacing the record in the database.

Could do more with it, if I had taken the time to code it.

--
Wes Groleau

There ain't no right wing,
there ain't no left wing.
There's only you and me and we just disagree.
(apologies to Jim Krueger)

This thread: