Denis Beauregard wrote: > On Wed, 18 Jul 2007 18:19:54 +0100, Peter J Seymour > <moz@pjsey.demon.co.uk> wrote in soc.genealogy.computing: > > >>Denis Beauregard wrote: >>... >> >>>The gain is because there are more and more tools to read XML >>>files and to handle this kind of data, event if it is not the >>>same as you use in genealogy. >>> >>>For example, a family is a set of persons where 2 persons have >>>a special role (dad and mom), and the other persons a more >>>general role (kids). You can say a car is a of pieces, or an >>>army regiment is a set of persons, etc. Then you can define >>>database operations like putting someone in the right box by >>>dragging that person, i.e. drag John to the dad's box or to the >>>captain's box, or drag the name of the cathedral to the relevant >>>box, etc. So, in theory, you could achieve many predefined >>>operations with the same code and getting some universal >>>genealogy software that will process the persons, but also the >>>events, the places, the groups (army, state employees, churchmen, >>>town elected, trades), the sources, etc. Then you get some kind >>>of very powerful genealogy system. >>> >> >>... >> >>>Denis >>> >> >>But that's not specific to XML? > > > What is specific to XML is the development of tools to handle this > kind of structures. GEDCOM can be developed by genealogists only. > XML can be developed by any programmer, so more development is > expected. > > > Denis > What I meant to add is that this is the sort of justification put forward by the LDS for their XML Gedcom. It didn't help them much. In my opinion, XML is a side issue. It may or may not have a role to play, depending on how people want to proceed. Appropriate data design is what counts and should come before implementation issues such as XML. Peter

In our last episode, <f7n62n$puv$1$8302bc10@news.demon.co.uk>, the lovely and talented Peter J Seymour broadcast on soc.genealogy.misc: > What I meant to add is that this is the sort of justification put > forward by the LDS for their XML Gedcom. It didn't help them much. In my > opinion, XML is a side issue. It may or may not have a role to play, > depending on how people want to proceed. Appropriate data design is what > counts and should come before implementation issues such as XML. The main advantage of XML is that namespaces will make it possible (and in well-behaved applications, necessary) for each product to define what it means by its tags. At present, that will make it possible for human developers to translate one format to another. The ultimate goal would seem to be to automate the business of translating from one namespace to another, so a very smart application that used a "spouse" tag could figure out how to handle input that used "HUSB" and "WIFE" without a human developer having to tell it. Of course, it never will be possible to get all the information from a more flexible format into a more restrictive one without losing some information. -- Lars Eighner <http://larseighner.com/> <http://myspace.com/larseighner> Countdown: 551 days to go. Owing to massive spam from googlegroups, I do not see most posts from there.