Bob Velke <bvelke@whollygenes.com> wrote: > Tom said: > > >GEDCOM is a perfectly good standard. It may be old, but the kilogram > >is an old standard and enjoys widespread modern use. > > For a lot of researchers, GEDCOM is far from "perfectly" anything. It misses few things so that the gedcom is a starting point for a serious standard. In other things, the fact that only 4 or 5 softwares respect it! > It's age is not the direct cause of its many weaknesses -- but the > length of time since a revision (and the lack of any prospect for > any) is a negative indicator for those of us who are affected. What good is it since the programs do not respect it ! It is not an update which would change that! -- Téhenne Saint-Denis de la Réunion Logiciel de généalogie ohmiGene (Mac & PC): http://www.nauze.com/ Comparatif Import-Export Gedcom : http://www.nauze.com/gedcom/ Digest du format Gedcom : http://www.nauze.com/format_gedcom/index.html

Téhenne said: >What good is it since the programs do not respect it ! > >It is not an update which would change that! It isn't as simple as that. All serious genealogy programs claim to import/export GEDCOM and many don't adhere to my interpretation of the specs. But some of those developers claim that my programs don't adhere to their interpretation of the specs. The fact that it is open to so many varied interpretations is one of its biggest weaknesses (e.g., the meaning of the FAMily structure, assumed marriage of parents, the sequence of undated children, precedence of names, handling of word-wrapping, unity of literally-identical sources, and on and on...). It is true that a few developers unapologetically manipulate the specs in a way which _cannot_ be defended as a difference in interpretation. While I think those cases are the minority, those developers would say that they've taken liberties out of a necessity to provide structured support for data types which are not directly addressed by GEDCOM (e.g, ambiguous parentage, shared non-spousal events, witnesses, research tasks, and on and on...). Some of us choose to not transfer those types of data with GEDCOM and others will manipulate it to fit GEDCOM's very-limited data model. But both of those are forms of data corruption, in my opinion. When added to all the problems of interpretation, the "it's better than nothing" argument becomes" it is better to corrupt it than to not transfer it" ... and that is at _least_ debatable. The cheerleading for GEDCOM on the grounds that there's nothing else (even if that were true), reminds me of when my kids complain when I won't drive them to the mall. They say, "How are we supposed to get there?" and I answer, "Who says that you're supposed to get there?" Over the years, there have been many revisions to GEDCOM which clarified some ambiguities, repaired some internal inconsistencies, and expanded its data model to account for some data types which its developers hadn't considered. That is, those revisions DID help to solidify the "standard." But it is still woefully insufficient on almost every scale - except among those researchers whose world view of genealogical data is limited to what PAF supports. In the their defense, that was GEDCOM developers' intended market and our application of it for a larger purpose was wishful thinking. It shouldn't surprise us that our current complaints with it will never be addressed. Bob Velke Wholly Genes, Inc. -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.503 / Virus Database: 269.15.31/1129 - Release Date: 11/13/2007 9:22 PM