GENBOX-L Archives

Archiver > GENBOX > 2005-02 > 1108512472


From: "Geir Thorud" <>
Subject: SV: [GENBOX] Import of place names from GEDCOM
Date: Wed, 16 Feb 2005 01:07:57 +0100
In-Reply-To: <005401c513a5$9c6937c0$6402a8c0@laptop>


Hi Wim,

I am aware that Gedcom does not support alternate names
since it has no separate structure or reference to tell which
alternate names belong to the same place. But, I do not
have an issue with that, and so should not Gedcom have - on
import.

Neither does it really matter how Legacy or TMG handles
this internally, what matters is the content of the Gedcom
file and how it is processed by Genbox.

Using your description of the matching algorithm I checked
the import of my file. The problem is perhaps not the matching
algorithm, but the way the name is stored. It looks like that
when a level is missing (,,) in the Gedcom, the the place is
stored as if the ",," was not there. For example with:

xxx,yyy,zzz,rrr,sss and
aaa, ,zzz,rrr,sss

aaa is stored at the same level as yyy, but should be stored at the
same level as xxx. Maybe the algorithm will work if this is fixed??

But there is still the issue about what to considder a Local site,
which will require some way to control the mapping. I would not like
to include Farm, Church etc in the name of the place since that will
not be "good" language.(I would have names like "Placename cottars farm",
which I would prefer to print as "the cottars farm Placename",
or for those understanding Scandinavian, I would prefer "husmannsplassen
Sorgen" rather than "Sorgen husmannsplassen". (husmannsplassen=the cottars
farm.) Lets hope for a Place_type field in a future Gedcom version.

A minor point re. TMG: If I edit a name in the master place list,
it is reflected in the event - at least in my 6.0 database.

Geir

> -----Opprinnelig melding-----
> Fra: wim prange [mailto:]
> Sendt: 15. februar 2005 22:31
> Til:
> Emne: Re: [GENBOX] Import of place names from GEDCOM
>
>
> Hi Geir,
>
> There are a couple of things Genbox will NOT do when importing
> place-records. First, Genbox will not create Placename-variants
> for places.
> The gedcom has no such thing for places so it would be wrong for GB to
> assume such a thing. First and foremost: Genbox should import the Place as
> is.
> Keep in mind, a placename-variant does not have to be another name for the
> place, but also a different higher level hierarchy.
>
> When importing, Genbox takes the entire name, including all higher levels,
> and starting at the lowest level, matches that against existing
> records. If
> the record exists and has the exact same hierarchy, it will take that
> record. But, as soon as there is a variation in the higher level
> (a missing
> one, or an extra level in between), it will create a separate placerecord
> (and another placenamerecord) for that place. GB then moves on to the next
> level in the hierarchy and runs this same routine from the start
> again: does
> this match another place+hierarchy already in the database or not.
> Afterwords the user should make the decision whether this is really a
> different place or only a placenamevariant.
>
> In regard to importing as Local Sites, GB is in my opinion a bit
> 'reluctant'
> to import places as Local Sites. If it can avoid it, it will preferably
> import it from City/Town and up. There is a provision for the English
> language in this: when the name has something like "Church",
> "Farm", "House"
> etc. in it, GB will use that and import the place as as a Local Site.
> Ofcourse, when the higher levels are allready there and the lowest is
> allready a City/Town, it then will proceed with giving it the level Local
> Site.
>
> My data came from Legacy which uses some sort of Places-table as
> well (I've
> never figured out how TMG had organised their places, it seemed to me that
> every mentioning of a place in events was a placerecord in
> itself: a change
> in the 'Masterrecord' did not reflect any change in an event itself).
> I assumed that I had a very clean placeslist in Legacy, with no
> 'inconsistencies', because I had devoted a lot of time and effort to it.
> After I reviewed my places in Genbox and went back to Legacy to check, I
> found that there were indeed some inconsistencies in the 'higher
> levels' in
> Legacy. It was only through Genbox that I found this out.
>
> Gr. Wim
>
>
>
>
>
> ----- Original Message -----
> From: "Geir Thorud" <>
> To: <>
> Sent: Tuesday, February 15, 2005 9:48 PM
> Subject: SV: [GENBOX] Import of place names from GEDCOM
>
>
> > Hi Anne,
> >
> > > Since I remeber the enormous amount of cleanup
> > > on places in TMG seems to me both program
> > > makes it necessary to work with it afterwards.
> >
> > I don't see a need for a cleanup if you have entered the
> > data properly in the first place. Could you give a gedcom
> > example describing the problem with TMG?
> >
> > > I think however it would be better if you gave concrete examples
> > > on what is in your Gedcom file, and how it is imported into
> > > Genbox and how you would expect it to be imported.
> >
> > I have included my first message posted 3 days ago, below.
> > It contains examples.
> >
> > Gr. Geir
> >
> > ------
> >
> > Hi,
> >
> > Genbox does not seem to import place hierarchies
> > properly when there are missing "levels" in the
> > hierarchy.
> >
> > For example:
> >
> > Mæhlum, Brøttum, Ringsaker, Hedmark - Mæhlum is imported as Local site.
> >
> > Tjerne, , Ringsaker, Hedmark - Tjerne is imported as City/Town
> > - note , , - missing level.
> >
> > I would expect both to be imported to the same level
> > in the hierarchy, i.e. as a Local site.
> >
> > ----
> >
> > Also, when I add country level in the Gedcom file,
> > for example
> >
> > Mæhlum, Brøttum, Ringsaker, Hedmark, Norway
> >
> > both of the two lover levels are imported as Local sites.
> > It would have been better to utilise more of the levels
> > in the Genbox hierarchy so that only one level in the
> > Gedcom file is mapped to Local site.
> >
> > In general more control over the mapping on import is
> > needed.
> >
> > Does anyone know if Genbox supports the FORM "Place_hierarchy"
> > structure in the Gedcom standard?
> >
> > Regards
> > Geir
> >
> >
> > ==== GENBOX Mailing List ====
> > To join this list, send an email to
> with the
> word "subscribe" as the subject line. Then email your messages to
> and they will appear on this list.
> >
> > ==============================
> > Search Family and Local Histories for stories about your family and the
> > areas they lived. Over 85 million names added in the last 12 months.
> > Learn more: http://www.ancestry.com/s13966/rd.ashx
> >
>
>
> ==== GENBOX Mailing List ====
> To join this list, send an email to
> with the word "subscribe" as the subject line. Then email your
> messages to and they will appear on this list.
>
> ==============================
> Search the US Census Collection. Over 140 million records added in the
> last 12 months. Largest online collection in the world. Learn
> more: http://www.ancestry.com/s13965/rd.ashx
>


This thread: