On Thu, Jul 12, 2001 at 09:31:10AM -0600, Don Allingham wrote:
> This is an interesting idea. Right now the concept of a place in gramps
> is very basic (just a text string). By changing to a place object, it
> would be much more powerful. I like the idea of being able to track
> places over time as names change.
>
> A couple of points to consider:
>
> 1) This would radically change the XML file format. Would everyone be
> willing to live with this? I could probably provide a means of
> translation (if gramps detected a place represented as a text string, it
> would create a new place object for it, so automagically upgrading the
> file). Compatibility with previous versions would be impacted - all
> place information would be lost if imported into an older version of
> gramps.
A different, though not really clean way, would be to still use place names
in person records and index the "place database" on these names instead of
on numbers like with persons and families. This would keep backwards
compatibility.
Another problem we have to think about is this: How do we keep track of the
place name used in the document describing some event. If somebody was born
in 'New Amsterdam', which was renamed 'New York' and in his marriage
certificate the name 'New York' is named we want to make sure that we
1) know its the same place and 2) keep (and show) the proper name for the
proper event so it corresponds to the original record.
> 2) Adding a new place would be more difficult, because instead of just
> entering a string, you would have to go through a selection process.
Not necessarily. The user interface could be the same as the when entering
surnames: You get a drop down box of all already known places and can choose
one or enter a new one. The added functionality could be hidden for the
casual user.
Jochen
--
Jochen Topf - jochen@... - http://www.remote.org/jochen/
|