TMG-L Archives

Archiver > TMG > 2011-03 > 1299629696

From: "Darrell A. Martin" <>
Subject: Re: [TMG] design question
Date: Tue, 08 Mar 2011 18:14:56 -0600
References: <24905F00F21F477D852F5DDD1D4DEA78@jfc> <>
In-Reply-To: <>


The problem with discussions about TMG and GEDCOM is that they usually
revolve around three propositions:

1. Here is what I wish GEDCOM would do.

2. GEDCOM does not do that.

3. GEDCOM is bad.

For the purposes for which it was created, GEDCOM is not merely
adequate; it works quite well. However, people found out it could *also*
be used as a tool to transfer data between genealogy programs. As those
programs get more sophisticated, the mismatch between what GEDCOM is --
and was intended to be -- and what it is often used for, grows larger.

GEDCOM was designed to transmit very specific kinds of information. It
continues to do that just as efficiently as it ever did. If the
information you want to transmit is not included in what GEDCOM was
designed to handle, then what you attempt may, or may not, work. That is
the fault of those who use it in a way not intended by its creators, not
of the specification itself.

"Fault" is not really a good description either, though. People want to
do something that is very reasonable: transfer data among genealogical
programs. The issue is not that GEDCOM was designed to meet that need,
and does a bad job of it -- GEDCOM was never designed to meet that need,
period. People use GEDCOM because it is the most capable tool that has
met with general acceptance. That isn't the same thing as saying it
works well, for purposes for which it was never intended.

Still, it *IS* better than nothing, by a long shot. Until there is a
reasonable alternative, it will keep being used as a transfer method.


This thread: