Family Historian User Group

I have adopted the convention of having 5 levels for all my Places, namely; village, town, city, county, country. This makes for very easy reviewing of my Place data.

However, in some cases I do not have a village name, or even a town name. This means that in many cases I have at level 1, and even at level 2.

My question is, should I maintain the hierarchy by using blanks (as I am doing now), or should I reduce the number of levels for those places where the lower level data is unknown or not applicable?

EG, if I don't have a village name, should I put town name into level 1 and only have 4 levels for that place? Or is it OK to have blank fields? In particular I want to ensure I can fully utilise the Google Maps feature.

Thanks

Eric

ID:4934

I maintain the same number of columns in all Place entries as you do, which may require fields sometimes.
This works better when using FH Tools > Work with Data > Places.
In the Places List window the columns all contain the same type of place (e.g. Place 1 = Town, Place 2 = County, Place 3 = Country).
Then sorting any column, by clicking on the column heading, gives sensible results, especially if Reverse Display Order is ticked.

I am not sure exactly what you mean by 'fully utilise the Google Maps feature'.
To link to Google Maps I use a URL such as http://maps.google.co.uk/maps?q=S5+7DQ where S5 7DQ is a modern postcode equivalent for the Place & Address.
Similar schemes can work for Grid Ref or Lat/Longitude.

FYI I use Place to hold the Registration District (post 1837) or Church Parish (pre 1837), and Address to hold the postal address, which sometimes is a different town &/or county.
For more details see my Forum topic Place and Address References at:
http://www.fhug.org.uk/cgi-bin/index.cg ... y&num=4706

Thanks again!

I'm glad you agree as I find it very satisfying to see all my Place data stacked up neatly in the Place List window.

My reference to Google Maps is because I am going to be using PHPGedView to display my data on the web, and it make use of Google Maps. I've uploaded a copy of my gedcom to an installation of PGV, and was expecting it to automatically locate all my places on the world map. Unfortunately it seems to be a little more complicated than that, although I have not yet had time to look into it in any detail (just 30 mins last night when I was already tired). Hopefully it won't be too hard to sort out.

When you say you put the postal address into the Address field, do you mean the whole address (house, street, town, postcode etc), which would in my case duplicate the town detail as this is already in my Place data? Or do you just mean the house and street, which is what I have done. Except, where I have it, I also include the parish as per the Census records, and the post code. Do you foresee this causing any problems?

Thanks

Eric

May I suggest you read through my Forum topic Place and Address References at:
http://www.fhug.org.uk/cgi-bin/index.cg ... y&num=4706
which goes into a lot more detail than I can repeat here.
It explains how I create a FH Source & Repository data structure for each unique Place & Address, and that is cited by every occurrence of that Place & Address.
This data structure holds all the details such as Multimedia that mention the Address and the Postcode, Latitude/Longitude, Grid Ref, and Google Maps URL for the Address.
This allows Reports and Websites to be created with Google Maps Hyperlinks for each Address purely from the FH GEDCOM data.

PhpGedView Google Maps
I had a quick look at PhpGedView and to make the Google Maps feature work it seems you have to manually enter a hierarchy of place details to match every Place and Address in your GEDCOM data.
These Google Maps place details are in another data structure outside the GEDCOM data and probably only usable in PhpGedView.
For this to work the PhpGedView advice is that all your Place & Address data must use modern place names otherwise Google Maps cannot find them.
Any original old place names it advises should be put in Notes.

Place & Postal Address
I use Place in a subtly different way than you.
My post-1837 Place fields hold the GRO Registration District, County & Country of the Event.
This is the District recorded in GRO BMD Index pages and on BMD Certificates.
Pre-1837 I use formal Church Parishes in a similar way.
My Address fields hold the postal address of house, street, town, etc, but not postcode, because that is in my cited Repository as explained above.
The GRO District name and Postal Village/Town name (and sometimes the County) are often different, especially around large cities like London.
For example:
Many old GRO Districts in London are in Middlesex County but the Postal Address is London.
Conversely, more recent GRO Districts are in London but the Postal Address is a home county such as Essex or Kent.

I hope this explains the differences in our approaches.

Thanks for the detailed response. I'll consider your advice, but I'm not sure I have the will to change all my addresses again. Right now I just want to get the PGV maps feature to work, and I'm struggling, but I guess that's a different forum.

Thanks again.

Hi,
A very interesting discussion and come at the right time as I am trying to go through the database and tidy things up after a couple of years learning to use FH (as compared to years using FTM)

A query has arisen with places like Barrow-in-Furness Census.
The Village is 'Barrow-in-Furness'
The Town is 'Barrow-in-Furness'
The Parish i 'Barrow-in-Furness'
The Sub District is 'Barrow-in-Furness'
The Reg District is 'Barrow-in-Furness'

Will tell me how to handle, in the structered manner, the above example as I have other places very similar, I know that this is an extreme condition.
Regards
John

I'm sure my approach could be criticized as 'too lazy', but I just enter what I think is needed to identify the Place in question unambiguously.  99% or more of my places are in England, so I decided not to add 'England' to the Place name, although I would use just 'England' if I didn't know a more precise Place name; (but I wouldn't bother to put ',,,England' in such cases, because I cannot see that it helps).  For John's example I would enter either 'Barrow-in-Furness, Lancashire' or 'Barrow-in-Furness, Lancashire, England', depending on the likely audience.  Obviously using extra commas to indicate missing parts of the Place name structure is needed if you wish to export the data to a spreadsheet with different place-name elements in separate columns.

The GEDCOM 5.5 standard has no limit on how many commas are used (just an upper limit of 120 characters in total), and FH supports up to 10 columns in its Place List Helper Dialog.

For the GRO registration period after 1837 I use the Registration District and County to formally identify the Place for each Fact, especially BMD and Census Records.
Then the Address field records the House/Street/Town postal address.

For example the GENUKI web site records the Barrow-in-Furness, Lanacashire Registration District at:
http://www.ukbmd.org.uk/genuki/reg/dist ... rness.html
and the starting point for the Districts data is:
http://www.genuki.org.uk/big/eng/
then click on the County, then Civil Registration, and then places in each.
Alternative, search for 'Barrow-in-Furness'.

So my Place entry for Barrow-in-Furness would be:
Barrow-in-Furness, Lancashire, England [GRO District]
and I always use the same number of commas so that FH Tools > Work with Data > Places sorting on columns and Reverse Display Order work rationally.
The England [GRO District] designation differentiates other places in England that are not GRO Districts.

As an addition to this discussion, when I open Tools/Work with Data/Places, I have the following problem:

I usually enter census places at Place, and the census address at Address. In some cases street numbers and names which I have entered at Address are listed under Places, and others are not.

What is FH's protocol for picking Address or Place

Roger

I am not aware of FH itself offering much advice on the use of Place and Address fields.

The GEDCOM 5.5 Specification has this to say:
PLACE_VALUE
'The jurisdictional name of the place where an event took place'
ADDRESS_STRUCTURE
'The address should be formed as it would appear on a mailing label'

The address structure includes sub-fields for Address Line 1 & 2, City, Country, Postcode, etc.
Therefore, I agree with you that Place should record the GRO District where a Census, Birth, Marriage, etc took place.
The Address should hold the house/street postal address.

An important use of Tools > Work with Data > Places is to help correlate your Place field contents.
If some Address details have mistakenly crept into your Place fields then Work with Data will help identify them.
Use the ability to sort the columns and Reverse Display Order to organise the details.
Select any Place entry and click the Report button to discover the Records that mention that Place.
The address details can then be moved from Place to Address until all the Place fields for the same GRO District correspond.
The Edit and Replace with buttons allow global changes to be made to the Place fields in multiple Records.

I will soon be posting a Knowledge Base How To for Create a Locations Database for Place & Address Details.

I have found out why Places sometimes shows the address and other times does not.
When I first started using FH, I entered the complete address at PLACE. I have since gone to the convention mentioned above, with the postal address at ADDRESS.

Ancestral Sources uses PLACE and ADDRESS. FH Reports put the Address as Additional Information.

Apparently Data/Places is just that - PLACES

I also use 3 levels in places, as per 'In the Places List window the columns all contain the same type of place (e.g. Place 1 = Town, Place 2 = County, Place 3 = Country). ' quoted above. I want to write a query that returns only the County, i.e. 'Place 2', but can't work out how to format the field. I had thought an export to a csv might work, but the whole place field gets put into a single column. I've tried searching the forums and help, to no avail, but may have missed something. Any help gratefully received. (ps. The Places repository looks interesting.)

You were close with the idea of saving the results to a CSV file.

If the only Column in the Result Set that contains commas (,) is the Place field then a simple text edit gets what you want.
Open the CSV file using a plain text editor such as Notepad or PSPad and use Edit > Replace to replace all string quotes (') with nothing, and save the resulting CSV file.
In other words delete all the string quotes surrounding the Place field, so that the commas within are now CSV value separators.
The three Place sub-fields each then appear in their own CSV column.