Re: [Gramps-users] Merging places?

[Espen B. <es...@gm...>]

2008/7/30 Martin, AA6E <mar...@gm...>:
> So I have 2,000 people in my system now, some arriving as imports from other
> workers.  I have a problem with "places":  There are *many* duplicate names
> and variants.   Trying to be tidy (and to clarify how to add places for new
> events), I would prefer to merge places when possible.  I know how to do it
> manually, but it's really tedious.  You have to track down each reference to
> a place and redirect it to the preferred place.
>
> Is there a tool to do this properly (i.e., recursively over all references
> in one go)?  Should there be?


Hi Martin!
Current status is that there is no automagic tool to merge places.
You have to to it one by one (or two by two ;) ).  I don't think an
automagic procedure to do this would be a great idea.  If place names
are registered differently, which criteria should be used to say -
yes, these two are actually the same.

Importing other people's work directly into your own database will for
sure give you this "problem" since different people have different
habits when it comes to registering names, places, sources and so on.
Personally I never import GEDCOM-files from other uses just because of
that.  Also, by importing from others, I don't really have any
knowledge of what information I actually imported.  What is deeply
lacking, what is totally wrong(!!!), and what is perfectly right?  The
best thing is then (my opinion) to create an empty database, import
the GEDCOM-file you got from some other person and create text-reports
(PDF or ODT or whatever) and simply start typing name by name what is
in the report.  Then you also will register places in the way you like
it. You can also verify information as you read them by consulting the
primary sources (or other thrustful sources).

So in this case, I wouldn't suggest an automagic recursive merging
algorithm to be implemented as I guess it would cause more mess than
what the users would like to see...


> In another sphere of life, I am a Python programmer, so "do it yourself"
> might be possible.  I wouldn't want to deprive anyone else of the challenge,
> however. ;-)

Then you could just dive into Mantis (http://bugs.gramps-project.org)
and create a patch to fix a bug or go through the feature requests and
see if some requests are easy to make.


> By the way, after some initial mis-starts on my part, I am finding Gramps to
> be a very neat system -- and it runs in Python on Linux!

Totally agree!!!
Good luck with your genealogy work.  It is addictive, and there is
(fortunately) no medicine preventing it :)


Espen