* Place/Address

[Johnyeates]

First of all may I wish everyone at the Forum a happy new year and thank them for their hard work during 2014.

Having just upgraded to V6 from V5 I have decided to have a clean sweep of all the old place names that I entered in the early days, around 1985/90 where I put everything into the place field and didn't use the Address Field.

I have been using the Place format of Local area, major town/Borough, County.

Now my query is that I cannot find in the FAQ is that Wandsworth in Surrey/London is in the Borough of Wandsworth, So do I enter the place as Wandsworth, Wandsworth, London?

My other query is Counties such as Surrey, Middlesex, Essex are now in London, so do I ignore the old county and just use London as the county, or do I have a fourth column for London, ie. Surrey, London?

Your ideas will be very enlightening.
John

[tatewise]

Happy New Year John.
I've moved this to General Usage where the Place/Address debate usually takes place.

Probably not too important about Wandsworth but could leave the first part blank:
, Wandsworth, London

Those home counties are not really in London. Parts of them may be in Greater London, but their postal address is still usually the county name, so stick with the county and it will probably geocode OK.

What I would strongly advise is a fourth Country column. You may know everywhere is in England but genealogists who inherit your data may not. Also, without the Country, it is more likely that geocoding will get it wrong where popular place names world wide could be chosen in error.

[jimlad68]

John, If you look through the forums you will notice that historically there has never been a "right" way to use PLACe and/or ADDRess. You need to decide how it best suites your needs and anyone else you might want to share your data with.

This FHUG Places and Addresses link might help:
http://www.fhug.org.uk/wiki/doku.php?id=glossary:places

Things have become more complicated recently with Geocoding, there are lots of links in this topic:
Best format of PLACe and ADDRess for Map Geocoding (11937)

[Johnyeates]

Thanks for the answers, I have a country column but I have only used it in the past for India Australia, USA etc.
All my modern entries have a country, it is just the hundreds or thousands of early entries that don't.
Sounds like I have a lot of editing to do.

[tatewise]

A combination of the new =TextPart() function, and Place Records, and a Plugin should allow it to be automated.

[jimlad68]

Not sure if this would work with the new v6 PLACE records, but a simple text editor (e.g. Notepad++) might do some of it more quickly than via FH.

Do a change for say , Preston (or whatever) to Preston , England, then find change repeat on required lines only. Although if you are up to it a Plugin like Mike suggestsmight be more efficient and accurate.

Whatever you do, try it first on a backup to make sure your methodology works.

[British Kiwi]

Hi
I have loads like this so what I have done is Wandsworth, London, Surrey, England. For newer records before it was absorbed by London I use Wandsworth, London, , England. By doing it this way I should be able to target certain counties when new records appear. Also where counties have changed I just use tge county of the time.

Good luck

[AdrianBruce]

For what it's worth...

One of the things to consider is what the data looks like in outputs, especially in a narrative report - and, let's face it, "Wandsworth, Wandsworth, London" doesn't look good does it?

I often use a 4-part place-name for the UK of "Local area, major town/Borough, County, Country". But there's no point in forcing the 4th element in if it doesn't add anything, so I often use just 3 parts. (I should probably put in a leading comma to line things up - e.g. ", Crewe, Cheshire, England" is probably better than "Crewe, Cheshire, England" as it will line the elements up and the report printing should take care of removing the leading comma. )

As for the London counties.... Rather than mess about entering things like "Camden, London, Middlesex, England" (I think that's right), I leave the county part as an empty part - thus: "Camden, London, , England". That's partly because as a Northerner I find it too tedious to look up the counties, when surely "London, , England" suffices? Also, logically, "London, Middlesex, England" and "London, Surrey, England" are 2 different places - and neither of them is "London, , England"!

Changing Counties over time? Many discussions on this.... Many people say: "Use the county applicable at the time". Well, yes - which county would that be? The administrative county? The ceremonial one? The historical one? The county corporate? And what about those many, many places in unitary authorities that aren't in a county at all these days?

My personal preference is to use the historical counties as drawn on the old maps, even for the current era.

This is all personal preference - think about it, stick to an idea and don't be afraid to use different naming schemes for different countries. Yes, I use the historical counties for England for all eras, while in America I tend to use the contemporary values.

[Johnyeates]

Thank you all for the answers, I think that I will use the current location name especially as others will be able to find it.
I have three location names for Bristol :- Gloucestershire, Avon and Gloucestershire, Avon but I have just found out that Bristol has been a county in its own right since 1373 so all of the foregoing is a load of bunkum and I will have to change them all to 'Bristol' as the county.
Thanks again
John

[TimTreeby]

Taken from the Bristol page of GENUKI http://www.genuki.org.uk/big/eng/GLS/Bristol/index.html

•Edward III made Bristol a "City and County" in 1373, although today, and in recent times past, it is not generally considered to be both. For genealogical purposes, and specifically in the International Genealogical Index (IGI) the City of Bristol is considered as part of the County of Gloucestershire. In contrast, the postal address prior to 1974, and the creation of the county of Avon was "Bristol, Somerset".

[AdrianBruce]

Bristol was a County Corporate so the city had some powers like a county. However for geographical purposes the rest of the world considered Bristol to be in Gloucestershire until the creation of the County of Avon.

If you describe Clifton for instance as Clifton, Bristol, County of Bristol, England that 3rd bit is useless. Clifton, Bristol, Gloucestershire, England is much more meaningful and is quite correct if you say that you're using the historical counties. It is also much more helpful when you're using a map.

[Johnyeates]

Just a quick note on modifying place names to a standard format.
After changing pages of them by hand, it occurred to me to use 'Find & Replace', this worked like a gem, why I didn't use it before I have no Idea, something I have not used before.
John

[tatewise]

You could also use the Search and Replace Plugin that can focus explicitly on say just Place fields/records to avoid changing anything else by accident, and it works in FH V5 too!

[jimlad68]

I tried that and works well, but unless I did something wrong, it just changed the "new PLACE" record and not those PLACe records associaited with events, hence when you out and back into FH, you get a double set of "new style" place records (old version recreated for the old event PLACes) and the ones changed previously.

Hope that makes sense just a rushed response before I fly.

[LornaCraig]

If you edit a place record the change is reflected in all fields using that place, no problem. It is the reverse that is not true: editing a particular place field will create a new place record (if one did not already exist for that place) but will not change the original place record or any other fields using it. That may be what caused the duplication you saw.

[jimlad68]

Thanks Lorna for that confirmation, so it looks like tatewise's Search and Replace with "Fact Place Fields" checked, only amends the "new PLACE" record and not those PLACe records associated with Facts.

I also could not get FHs V6 new Find and Replace to do it either. So my only quick answer seems to be to use something like Notepad++ with the Gedcom file to multiple edit the Fact PLACes, not a big hasle, but I am always wary of directly editing the Gedcom direct.

Does anyone know of another way to multiple change/replace PLACe records associated with Facts,, perhaps I have not been doing it right with the above 2.

[tatewise]

Jim, I am not sure I follow either your logic or your terminology.

I don't understand what you mean by 'only amends the "new PLACE" record'.
PLACe field values and their linked _PLACe records will always change in tandem.
This is true even if you edit the GEDCOM directly, because as soon as it is opened by FH any missing _PLACe records get created.

I also don't know what you mean by 'PLACe records associated with Facts'.
PLACe records are NOT associated with Facts. (Strictly speaking they are actually _PLACe records.)
_PLACe records are linked to PLACe fields such that each PLACe field with the same value is linked to just one _PLACe record that has its Name equal to that value. It is that unique value that links them together.

I will double-check tomorrow, but I believe that if the Search and Replace Plugin changes a set of PLACe fields, then a totally consistent set of changes will apply to their linked _PLACe records.

However, residual _PLACe records with Name values no longer used by any PLACe fields may result from any of the above changes, but they will have 0 links to PLACe fields, and can easily be deleted as necessary.
This preservation of unused _PLACe records was requested in order to maintain a _PLACe records database even if some are not used by any PLACe fields.

[LornaCraig]

Jim,

Perhaps I am misunderstanding what you are trying to do.

My understanding is that there is only one kind of Place record (what you call an "new PLACE" record) and the Fact Place Fields point/link to a Place record.

If you use Tools>Work with Data> Places and edit a Place record there, all instances of that place in the Gedcom will change, and I think that is the global change you want.

[tatewise]

Jim, Lorna's description is consistent with my understanding, except to say that PLACe fields exist in more than just Facts, and can be found in SOURce records, HEADer record, and LDS ordinances.

Lorna, I think Jim wants to make more global changes than to just one _PLACe record. For example, correct the spelling of a County used in many _PLACe records and PLACe fields.

[jimlad68]

Mike, Lorna, many thanks for those observations which I will study and experiment with later and respond when I next have internet connectivity. I must try harder with my "terminology"!!

[tatewise]

I have reviewed how Search and Replace operates for Place name changes, and it does offer all the necessary options, but it is important to choose the right ones for your objectives.

Here are a couple of typical scenarios:

(1) Global Change
The objective is to make global changes to existing Place names, but without altering the links between PLACe fields and _PLACe records, and retaining all _PLACe record details such as Lat/Longitude, Notes, Media, etc.

(a) In the Major Options tab set Search Scope: Place Records (_PLAC) and tick Fact Place Fields but clear all other Basic Filters.
OR
(b) Leave Search Scope: All Records & Events/Attributes and in Extra Filters tab clear all filters except under Distinctive Fields select Record Title/Names for Place Name (_PLAC.TEXT).

This technique should never result in residual unused _PLACe records with 0 links.

(2) Selective Change
The objective is to make selective changes to some Place names, and for new PLACe field values to create new blank _PLACe records, but leave all other PLACe fields and _PLACe records unaltered.

(a) Leave Search Scope: All Records & Events/Attributes and tick Fact Place Fields but clear all other Basic Filters. BUT, when the Data Reference in replacement dialogue reaches the first _PLAC Record Id: [*] Data Ref: _PLAC.TEXT[/b] click Cancel to prevent altering any existing _PLACe records. The snag is if there are a lot of replacements to Confirm, then it is tedious stepping through to the first _PLACe record, but if the changes are selective you will probably need to Confirm each one anyway.
OR
(b) Set Search Scope: Individual Records (INDI) or Family Records (FAM) or any other Search Mode to narrow the selection, and tick Fact Place Fields but clear all other Basic Filters. Then repeat with other Search Scope: filters until all required PLACe fields are changed. This approach may allow bulk changes by unticking Confirm every item found.

This technique may leave residual unused _PLACe records with 0 links, that may need to be Merged with new blank _PLACe records to migrate the details such as Lat/Longitude, Media, etc.

(3) Find and Replace
Do NOT use FH Edit > Find and Replace for Places if your objective is (1) above, because it will behave like (2) above, and create new blank _PLACe records for every new PLACe field value.

[LornaCraig]

Mike said:

I think Jim wants to make more global changes than to just one _PLACe record. For example, correct the spelling of a County used in many _PLACe records and PLACe fields.

Jim said:

I also could not get FHs V6 new Find and Replace to do it either.

I have just tried this and it does appear to work, but perhaps there is some other snag I am not aware of?

Starting with two places, ‘Town1, CountyA’ and ‘Town2, CountyA’ use Edit>Find and Replace to search for CountyA and replace with CountyB.

Un-tick all Look in options except Places.

The result will be two places called ‘Town1, CountyB’ and ‘Town2, CountyB’. There are no residual unused CountyA places left. All fields which used the CountyA places now use the CountyB places.

It also works if a Place record for ‘Town2, CountyB’ already existed: its used count is increased, and the CountyA places are removed.

Forgive me if I am still missing something - (entirely possible!)

(Edit: Mikes's post came in while I was writing this, so this may now be irrelevant - sorry)

[tatewise]

Lorna, my experience of Find and Replace is different.

Taking exactly your example it resulted in four Place records:
Town1, CountyA with Links = 0
Town2, CountyA with Links = 0
Town1, CountyB with Links = 1
Town2, CountyB with Links = 1

First two are residual unused records with their original Lat/Longitude, Notes, Media, etc.
Second two are the new blank records.

So if your objective was to simply correct a spelling in the County name, you now have to Merge the pairs of records to migrate all the details from old records to new records..

[LornaCraig]

Mike,

In my tests I did not have any Lat/Long, Notes or Media – just the two Place records with names. Repeating my test with these I still (superficially) see the same result, and after doing the Find and Replace, if I look at the Places tab of the Records window while I use Edit > undo and Edit > redo I can watch the names of the two Places switching repeatedly between the CountyA and CountyB versions. There are no extra residual unused records visible at any stage.

However I now realise that this is because FH does in fact automatically delete unused records if they are ‘empty’, i.e. if they have a name and nothing else. So in fact a new Place record is being created and the old one deleted, but it looks as if the Place name is just being edited. Now that I look at the Record IDs of the Place records this is obvious, of course!

We live and learn... 

[LornaCraig]

Just to add: this behaviour in Edit> Find and Replace is inconsistent with the rest of FH. In all other circumstances that I have tried, FH retains even ‘empty’ unused Place records which have only a name.

I guess the deletion by Find and Replace is a bug, but it’s a fairly obscure one that it not likely to cause problems (other than my confusion, above!)