Place types in v4.1

classic Classic list List threaded Threaded
29 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Place types in v4.1

Nick Hall
Devs,

The pre-defined place types in v4.1 can be divided into three groups.

1. Settlements

City * - A large town, normally having a cathedral, local government
function, or established by a charter.
Town - A densely populated urban area.
Village - A small community usually consisting of a church and local
amenities such as a post office, village hall, pub etc...
Hamlet - A small group of houses.

2. Address

Locality * - A local area or a suburb of a town or city.
Neighbourhood - An area immediately surrounding a location.
Street * - A road usually lined with buildings.
Building
Farm

Do we need to add any more?  E.g. Cemetery.  Do we need Neighbourhood?

2. Administrative Areas

The following are administrative areas arranged in approximate size order:

Country *
State *
Province
Region
County *
Department
Borough
District
Municipality
Parish *

The 4 marked with (*) are in previous versions of Gramps.

Again, do we need them all?  Anything missing?  E.g. Territory.

Any comments?  What problems does this cause for translators?

Regards,


Nick.


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

Charlie Wilson
Hello. I'm new to the neighborhood. ;)

I personally would use Locality and Neighborhood. There is a difference but I suspect many would use them interchangeably.

Locality - christian, jewish, asian american, working-class, beer making. I reserve it for types of people, things and ,activities going on in an area. Neighborhood (Seattle) - Belltown, Queen Anne, Ballard, ...

Administrative Areas - Parcel? Many deed and other documents will refer all the way down to the parcel number.

Thanks,
Charlie


On Sun, May 18, 2014 at 11:24 AM, Nick Hall <[hidden email]> wrote:
Devs,

The pre-defined place types in v4.1 can be divided into three groups.

1. Settlements

City * - A large town, normally having a cathedral, local government
function, or established by a charter.
Town - A densely populated urban area.
Village - A small community usually consisting of a church and local
amenities such as a post office, village hall, pub etc...
Hamlet - A small group of houses.

2. Address

Locality * - A local area or a suburb of a town or city.
Neighbourhood - An area immediately surrounding a location.
Street * - A road usually lined with buildings.
Building
Farm

Do we need to add any more?  E.g. Cemetery.  Do we need Neighbourhood?

2. Administrative Areas

The following are administrative areas arranged in approximate size order:

Country *
State *
Province
Region
County *
Department
Borough
District
Municipality
Parish *

The 4 marked with (*) are in previous versions of Gramps.

Again, do we need them all?  Anything missing?  E.g. Territory.

Any comments?  What problems does this cause for translators?

Regards,


Nick.


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

enno
In reply to this post by Nick Hall
Nick,
> 1. Settlements
>
> City * - A large town, normally having a cathedral, local government
> function, or established by a charter.
> Town - A densely populated urban area.
> Village - A small community usually consisting of a church and local
> amenities such as a post office, village hall, pub etc...
> Hamlet - A small group of houses.
In general, I don't care much about these types, and here in The
Netherlands, your list will be difficult to translate, because we mostly
rely on stad (city) and dorp (village), where city is not defined by
size, but by status. In the past, cities were populated places that got
grants to build a (city) wall, where villages were not allowed to do
that. Towns do not exist as a separate class, making that term hard to
translate. I can probably create a valid translation for Hamlet though.

My personal preference is to record a place of birth (geboorteplaats),
because that's how it appears in my passport, and in records that I
find. Here, Place is a neutral term that can refer to a city or village,
or whatever you like. Point is, that I don't care, and I don't want to
be bothered with it either.

In 3.4 it means that I prefer to translate City to Plaats, because
that's what it is.
> 2. Address
>
> Locality * - A local area or a suburb of a town or city.
> Neighbourhood - An area immediately surrounding a location.
> Street * - A road usually lined with buildings.
> Building
> Farm
>
> Do we need to add any more?  E.g. Cemetery.  Do we need Neighbourhood?
I support Cemetery, and Parcel, mentioned in another message. And I
object to the translation of Locality to Plaats as is the case in 3.4.
That's why I run a hacked Gramps here, where Locality is Wijk.

> 2. Administrative Areas
>
> The following are administrative areas arranged in approximate size order:
>
> Country *
> State *
> Province
> Region
> County *
> Department
> Borough
> District
> Municipality
> Parish *
>
> The 4 marked with (*) are in previous versions of Gramps.
>
> Again, do we need them all?  Anything missing?  E.g. Territory.
>
> Any comments?  What problems does this cause for translators?
I don't see translation problems for Dutch here, but I will probably
just use a few, even for ancestors from England, France, or Germany.

I checked some other resources on the web, and found that GedcomX does
not have place type definitions yet, and the FamilySearch catalog is
quite chaotic in a way. I do know a developer who is quite active in the
are though, and that's Dallan Quass, creator of WeRelate.org. On his
site, you can check

http://www.werelate.org/wiki/Place:Bermondsey%2C_London%2C_England

and

http://www.werelate.org/wiki/Place:Spenge%2C_Westfalen%2C_Preu%C3%9Fen%2C_Germany

for examples of how things can be done. I mention both, because I have
ancestors there, and they refer to different resources, which may both
be of interest for us. The 1st is wikipedia, the 2nd Getty, and I think
both are more consistent than the FamilySearch catalog.

When you try the search on Getty, you will see that they invariably
refer to city/town/village as inhabited place, and I really like that.
They also refer to Bermondsey as a neighborhood, in a borough, in London
(inhabited place), which is exactly how it is today.

For Spenge, Getty leaves out the District (Kreis), which is mentioned on
wikipedia:

http://en.wikipedia.org/wiki/Spenge

This part is relevant, because it's used locally, unlike the national
district mentioned by Getty, both in civil use, and for the church district.

It's a long story, but to summarize, I prefer the Getty system of
inhabited place, and wikipedia where Getty's incomplete.

regards,

Enno


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

ACProctor
Ah, that old subject... Thinking "sideways", why does Gramps need a single
set of types that applies to every country? Let me explain...

Each country has its own divisions/subdivisions, whether geographical or
administrative. The list in this thread already excludes some, e.g. Townland
(Ireland) and Township (several countries). My approach was to collect a
rather large set of such types which included as many countries as I knew
of, and then allow a configurable subset to be used for each country. That
means I can still use a single common vocabulary, but each country is not
forced to utilise every single possibility, e.g. "State" in the UK doesn't
mean anything. "Province" exists in both UK (although rarely used), Ireland,
and Netherlands but not in the US. However, a State cannot be equated with a
Province since they are quite different in concept.

    Tony Proctor

----- Original Message -----
From: "Enno Borgsteede" <[hidden email]>
To: <[hidden email]>
Sent: Monday, May 19, 2014 1:56 PM
Subject: Re: [Gramps-devel] Place types in v4.1


> Nick,
>> 1. Settlements
>>
>> City * - A large town, normally having a cathedral, local government
>> function, or established by a charter.
>> Town - A densely populated urban area.
>> Village - A small community usually consisting of a church and local
>> amenities such as a post office, village hall, pub etc...
>> Hamlet - A small group of houses.
> In general, I don't care much about these types, and here in The
> Netherlands, your list will be difficult to translate, because we mostly
> rely on stad (city) and dorp (village), where city is not defined by
> size, but by status. In the past, cities were populated places that got
> grants to build a (city) wall, where villages were not allowed to do
> that. Towns do not exist as a separate class, making that term hard to
> translate. I can probably create a valid translation for Hamlet though.
>
> My personal preference is to record a place of birth (geboorteplaats),
> because that's how it appears in my passport, and in records that I
> find. Here, Place is a neutral term that can refer to a city or village,
> or whatever you like. Point is, that I don't care, and I don't want to
> be bothered with it either.
>
> In 3.4 it means that I prefer to translate City to Plaats, because
> that's what it is.
>> 2. Address
>>
>> Locality * - A local area or a suburb of a town or city.
>> Neighbourhood - An area immediately surrounding a location.
>> Street * - A road usually lined with buildings.
>> Building
>> Farm
>>
>> Do we need to add any more?  E.g. Cemetery.  Do we need Neighbourhood?
> I support Cemetery, and Parcel, mentioned in another message. And I
> object to the translation of Locality to Plaats as is the case in 3.4.
> That's why I run a hacked Gramps here, where Locality is Wijk.
>> 2. Administrative Areas
>>
>> The following are administrative areas arranged in approximate size
>> order:
>>
>> Country *
>> State *
>> Province
>> Region
>> County *
>> Department
>> Borough
>> District
>> Municipality
>> Parish *
>>
>> The 4 marked with (*) are in previous versions of Gramps.
>>
>> Again, do we need them all?  Anything missing?  E.g. Territory.
>>
>> Any comments?  What problems does this cause for translators?
> I don't see translation problems for Dutch here, but I will probably
> just use a few, even for ancestors from England, France, or Germany.
>
> I checked some other resources on the web, and found that GedcomX does
> not have place type definitions yet, and the FamilySearch catalog is
> quite chaotic in a way. I do know a developer who is quite active in the
> are though, and that's Dallan Quass, creator of WeRelate.org. On his
> site, you can check
>
> http://www.werelate.org/wiki/Place:Bermondsey%2C_London%2C_England
>
> and
>
> http://www.werelate.org/wiki/Place:Spenge%2C_Westfalen%2C_Preu%C3%9Fen%2C_Germany
>
> for examples of how things can be done. I mention both, because I have
> ancestors there, and they refer to different resources, which may both
> be of interest for us. The 1st is wikipedia, the 2nd Getty, and I think
> both are more consistent than the FamilySearch catalog.
>
> When you try the search on Getty, you will see that they invariably
> refer to city/town/village as inhabited place, and I really like that.
> They also refer to Bermondsey as a neighborhood, in a borough, in London
> (inhabited place), which is exactly how it is today.
>
> For Spenge, Getty leaves out the District (Kreis), which is mentioned on
> wikipedia:
>
> http://en.wikipedia.org/wiki/Spenge
>
> This part is relevant, because it's used locally, unlike the national
> district mentioned by Getty, both in civil use, and for the church
> district.
>
> It's a long story, but to summarize, I prefer the Getty system of
> inhabited place, and wikipedia where Getty's incomplete.
>
> regards,
>
> Enno
>
>
> ------------------------------------------------------------------------------
> "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
> Instantly run your Selenium tests across 300+ browser/OS combos.
> Get unparalleled scalability from the best Selenium testing platform
> available
> Simple to use. Nothing to install. Get started now for free."
> http://p.sf.net/sfu/SauceLabs
> _______________________________________________
> Gramps-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gramps-devel 


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

enno
Tony,

> Ah, that old subject... Thinking "sideways", why does Gramps need a
> single set of types that applies to every country? Let me explain...
>
> Each country has its own divisions/subdivisions, whether geographical
> or administrative. The list in this thread already excludes some, e.g.
> Townland (Ireland) and Township (several countries). My approach was
> to collect a rather large set of such types which included as many
> countries as I knew of, and then allow a configurable subset to be
> used for each country. That means I can still use a single common
> vocabulary, but each country is not forced to utilise every single
> possibility, e.g. "State" in the UK doesn't mean anything. "Province"
> exists in both UK (although rarely used), Ireland, and Netherlands but
> not in the US. However, a State cannot be equated with a Province
> since they are quite different in concept.
True, but here lies a problem. Types are nice for developers, but users
don't see types, but words, mostly in their own language, and some types
are simply impossible to translate. That's why the combination of city,
town, and village, is completely unacceptable for me. Official documents
refer to the neutral place (plaats), or one of 2 types: city (stad), or
village (dorp). Towns do not exist in The Netherlands, and since we have
no word for those, introducing those in Gramps will inevitably lead to
duplicates in the translated type table, just like it already does for
baptism and christening in my dictionary.

State is easier to handle, because confusion can be avoided by adding
context. We refer to States in the US with the Dutch word Staat, but we
also refer to our own country as the Staat der Nederlanden, where Staat
means Country, Land in Dutch. And in Germany, they use Land for State.
Got that?

Interestingly, I saw no types for Kingdom or Republic yet. What would
you think of those?

regards,

Enno


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

ACProctor
Kingdom and Republic are more at the "nation" level, Enno, and often
included with "country". I know this isn't always true, though, and a good
example is the UK (United Kingdom), which is technically a "sovereign state"
built of distinct countries, plus one dependent territory (N. Ireland). I
*never*  include UK in any of my own places as it is unnecessary and a
moving target historically, just as saying "Europe" would be unnecessary.

My only criticism of what you said would be "...but users don't see types,
but words...". Those translations don't have to generate a single word, and
I would fully expect many to require a short descriptive phrase. As an
example of this, consider the US which is widely deemed to have 50 states
(and historically less than that). However, Washington DC isn't in any of
those States, and is classed as a "federal district" (DC = District of
Columbia). I have seen people recommend that they just enter "Washington DC"
as a pseudo-State, and or even a pseudo-County with an empty State, but
they're technically wrong approaches that are necessary only because the
associated software was too limited. Having a US-specific type whose
end-user description would be "federal district" doesn't sound a big problem
to me.

    Tony Proctor

----- Original Message -----
From: "Enno Borgsteede" <[hidden email]>
To: "Tony Proctor" <[hidden email]>; <[hidden email]>
Sent: Monday, May 19, 2014 2:58 PM
Subject: Re: [Gramps-devel] Place types in v4.1


> Tony,
>> Ah, that old subject... Thinking "sideways", why does Gramps need a
>> single set of types that applies to every country? Let me explain...
>>
>> Each country has its own divisions/subdivisions, whether geographical or
>> administrative. The list in this thread already excludes some, e.g.
>> Townland (Ireland) and Township (several countries). My approach was to
>> collect a rather large set of such types which included as many countries
>> as I knew of, and then allow a configurable subset to be used for each
>> country. That means I can still use a single common vocabulary, but each
>> country is not forced to utilise every single possibility, e.g. "State"
>> in the UK doesn't mean anything. "Province" exists in both UK (although
>> rarely used), Ireland, and Netherlands but not in the US. However, a
>> State cannot be equated with a Province since they are quite different in
>> concept.
> True, but here lies a problem. Types are nice for developers, but users
> don't see types, but words, mostly in their own language, and some types
> are simply impossible to translate. That's why the combination of city,
> town, and village, is completely unacceptable for me. Official documents
> refer to the neutral place (plaats), or one of 2 types: city (stad), or
> village (dorp). Towns do not exist in The Netherlands, and since we have
> no word for those, introducing those in Gramps will inevitably lead to
> duplicates in the translated type table, just like it already does for
> baptism and christening in my dictionary.
>
> State is easier to handle, because confusion can be avoided by adding
> context. We refer to States in the US with the Dutch word Staat, but we
> also refer to our own country as the Staat der Nederlanden, where Staat
> means Country, Land in Dutch. And in Germany, they use Land for State. Got
> that?
>
> Interestingly, I saw no types for Kingdom or Republic yet. What would you
> think of those?
>
> regards,
>
> Enno
>


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

jerome
In reply to this post by Nick Hall
Nick,

About translations, it seems that we are creating custom types!
So, maybe a minor GrampsType handling issue?
We already had such issues in the past. 
I will look at this.

Note, Gramps will be available to handle the new french "province" level!
Really great!

Current top 'local' areas:

Most archives are on the country level (children of current french provinces).
Most of my ancestors could be grouped into such a mega-province.


Thank you.
Jérôme


Le dim. 18 mai 2014 at 20:24, Nick Hall <[hidden email]> a écrit :
Devs, The pre-defined place types in v4.1 can be divided into three groups. 1. Settlements City * - A large town, normally having a cathedral, local government function, or established by a charter. Town - A densely populated urban area. Village - A small community usually consisting of a church and local amenities such as a post office, village hall, pub etc... Hamlet - A small group of houses. 2. Address Locality * - A local area or a suburb of a town or city. Neighbourhood - An area immediately surrounding a location. Street * - A road usually lined with buildings. Building Farm Do we need to add any more? E.g. Cemetery. Do we need Neighbourhood? 2. Administrative Areas The following are administrative areas arranged in approximate size order: Country * State * Province Region County * Department Borough District Municipality Parish * The 4 marked with (*) are in previous versions of Gramps. Again, do we need them all? Anything missing? E.g. Territory. Any comments? What problems does this cause for translators? Regards, Nick. ------------------------------------------------------------------------------ "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE Instantly run your Selenium tests across 300+ browser/OS combos. Get unparalleled scalability from the best Selenium testing platform available Simple to use. Nothing to install. Get started now for free." http://p.sf.net/sfu/SauceLabs _______________________________________________ Gramps-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gramps-devel

------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

enno
In reply to this post by ACProctor
Tony,
> Kingdom and Republic are more at the "nation" level, Enno, and often
> included with "country". I know this isn't always true, though, and a
> good example is the UK (United Kingdom), which is technically a
> "sovereign state" built of distinct countries, plus one dependent
> territory (N. Ireland). I *never*  include UK in any of my own places
> as it is unnecessary and a moving target historically, just as saying
> "Europe" would be unnecessary.
I know that, but I brought it up deliberately, because there is a
parallel with place as it shows in my passport. That passport refers to
nationality (nation) and a place, and nothing else, location wise,
meaning that it doesn't tell whether my place of birth is a city, town,
village, whatever.

> My only criticism of what you said would be "...but users don't see
> types, but words...". Those translations don't have to generate a
> single word, and I would fully expect many to require a short
> descriptive phrase. As an example of this, consider the US which is
> widely deemed to have 50 states (and historically less than that).
> However, Washington DC isn't in any of those States, and is classed as
> a "federal district" (DC = District of Columbia). I have seen people
> recommend that they just enter "Washington DC" as a pseudo-State, and
> or even a pseudo-County with an empty State, but they're technically
> wrong approaches that are necessary only because the associated
> software was too limited. Having a US-specific type whose end-user
> description would be "federal district" doesn't sound a big problem to
> me.
I know what you mean, but as a user, I like to see simple menu's, and
that includes short lists of types. So, when I enter a location in The
Netherlands, I like to see only those types that are relevant, which
means that at that time, I'm absolutely not interested in counties,
federal districts, states, etc.

But OTOH, when I do enter a location in England, France, Germany, the
U.A.E., the U.S.A., it does help if the type list adapts, which is sort
of equivalent to my wish for easy citation templates, which are also
still not defined in GedcomX, btw, nor by the FHISO, or what's left of
that. :-)

The citation templates didn't make it into Gramps 4.1, and I prefer that
place types CAN be used. But they will only work for me, when they
follow a simple scheme, which exports to GEDCOM too, and maybe that is a
bit template like.

regards,

Enno


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

ACProctor
The adaptive-list is what I meant in my earlier post Enno when I said "My
approach was ..., and then allow a configurable subset to be used for each
country". In other words, we're agreeing Enno  :-)

    Tony Proctor

----- Original Message -----
From: "Enno Borgsteede" <[hidden email]>
To: "Tony Proctor" <[hidden email]>; <[hidden email]>
Sent: Monday, May 19, 2014 7:28 PM
Subject: Re: [Gramps-devel] Place types in v4.1


> Tony,
>> Kingdom and Republic are more at the "nation" level, Enno, and often
>> included with "country". I know this isn't always true, though, and a
>> good example is the UK (United Kingdom), which is technically a
>> "sovereign state" built of distinct countries, plus one dependent
>> territory (N. Ireland). I *never*  include UK in any of my own places as
>> it is unnecessary and a moving target historically, just as saying
>> "Europe" would be unnecessary.
> I know that, but I brought it up deliberately, because there is a parallel
> with place as it shows in my passport. That passport refers to nationality
> (nation) and a place, and nothing else, location wise, meaning that it
> doesn't tell whether my place of birth is a city, town, village, whatever.
>> My only criticism of what you said would be "...but users don't see
>> types, but words...". Those translations don't have to generate a single
>> word, and I would fully expect many to require a short descriptive
>> phrase. As an example of this, consider the US which is widely deemed to
>> have 50 states (and historically less than that). However, Washington DC
>> isn't in any of those States, and is classed as a "federal district" (DC
>> = District of Columbia). I have seen people recommend that they just
>> enter "Washington DC" as a pseudo-State, and or even a pseudo-County with
>> an empty State, but they're technically wrong approaches that are
>> necessary only because the associated software was too limited. Having a
>> US-specific type whose end-user description would be "federal district"
>> doesn't sound a big problem to me.
> I know what you mean, but as a user, I like to see simple menu's, and that
> includes short lists of types. So, when I enter a location in The
> Netherlands, I like to see only those types that are relevant, which means
> that at that time, I'm absolutely not interested in counties, federal
> districts, states, etc.
>
> But OTOH, when I do enter a location in England, France, Germany, the
> U.A.E., the U.S.A., it does help if the type list adapts, which is sort of
> equivalent to my wish for easy citation templates, which are also still
> not defined in GedcomX, btw, nor by the FHISO, or what's left of that. :-)
>
> The citation templates didn't make it into Gramps 4.1, and I prefer that
> place types CAN be used. But they will only work for me, when they follow
> a simple scheme, which exports to GEDCOM too, and maybe that is a bit
> template like.
>
> regards,
>
> Enno
>


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

enno
Tony,
> The adaptive-list is what I meant in my earlier post Enno when I said
> "My approach was ..., and then allow a configurable subset to be used
> for each country". In other words, we're agreeing Enno  :-)
Good ... and to be honest, I must say that we have an easy job,
theorizing ...

Any ideas about that adaptive list? One that also works in the past,
when there were still kingdoms in Germany?

cheers,

Enno


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

ACProctor
I would prefer to select a region name from a list first. For instance,
'England' or 'Netherlands', and then have it present a form with the
appropriate fields in. If there were some important variation (e.g.
'Medieval England' in order to include "Shires") then it could be a separate
entry on that list. This is how I do it at the moment, and it's nice and
simple, but some people might want configuration options to control the
region names that they're presented with, just to make the UI more slick.

    Tony Proctor

----- Original Message -----
From: "Enno Borgsteede" <[hidden email]>
To: "Tony Proctor" <[hidden email]>
Cc: <[hidden email]>
Sent: Monday, May 19, 2014 7:43 PM
Subject: Re: [Gramps-devel] Place types in v4.1


> Tony,
>> The adaptive-list is what I meant in my earlier post Enno when I said "My
>> approach was ..., and then allow a configurable subset to be used for
>> each country". In other words, we're agreeing Enno  :-)
> Good ... and to be honest, I must say that we have an easy job, theorizing
> ...
>
> Any ideas about that adaptive list? One that also works in the past, when
> there were still kingdoms in Germany?
>
> cheers,
>
> Enno
>


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

Nick Hall
In reply to this post by enno
On 19/05/14 13:56, Enno Borgsteede wrote:
> When you try the search on Getty, you will see that they invariably
> refer to city/town/village as inhabited place, and I really like that.
> They also refer to Bermondsey as a neighborhood, in a borough, in London
> (inhabited place), which is exactly how it is today.

Getty has many place types including city, town, village and inhabited
place.  It allows more than one type per place.

For example, Canterbury is listed as both a city and inhabited place.

Bermondsey is listed as a historic borough, and also a neighbourhood
within a modern borough.

Nick.




------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

Nick Hall
In reply to this post by ACProctor
On 19/05/14 14:12, Tony Proctor wrote:
> Ah, that old subject... Thinking "sideways", why does Gramps need a single
> set of types that applies to every country? Let me explain...

No.  A single list of types was easiest to code.  The user can add to it
by defining custom types.


> Each country has its own divisions/subdivisions, whether geographical or
> administrative. The list in this thread already excludes some, e.g. Townland
> (Ireland) and Township (several countries). My approach was to collect a
> rather large set of such types which included as many countries as I knew
> of, and then allow a configurable subset to be used for each country. That
> means I can still use a single common vocabulary, but each country is not
> forced to utilise every single possibility, e.g. "State" in the UK doesn't
> mean anything. "Province" exists in both UK (although rarely used), Ireland,
> and Netherlands but not in the US. However, a State cannot be equated with a
> Province since they are quite different in concept.

I started with a small list of seven items.  These were the fields in
the previous version of Gramps.

One option would only have these 7 pre-defined items and let the user
add custom types when necessary.  I decided to add some extra ones.  My
test data consisted of English, American and French places.  This is why
I added Region, Department and Borough.  Users requested Municipality
and Farm.

I soon realised that the list was getting large.  If I split out
building types it would get even larger.

Letting the user define subsets of the list for different countries is a
good idea.

Nick.


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

Nick Hall
In reply to this post by ACProctor
On 19/05/14 16:31, Tony Proctor wrote:
Kingdom and Republic are more at the "nation" level, Enno, and often 
included with "country". I know this isn't always true, though, and a good 
example is the UK (United Kingdom), which is technically a "sovereign state" 
built of distinct countries, plus one dependent territory (N. Ireland). I 
*never*  include UK in any of my own places as it is unnecessary and a 
moving target historically, just as saying "Europe" would be unnecessary.

Gramps uses the country field for nation state.  I have kept it like that with a single pre-defined "Country" type.  I didn't see much point in introducing other top level types.  There is nothing to stop a user from defining custom types though.

I don't use "UK" in my place data either.  At the moment I just leave the state field blank.  For London I put "London" in the county field and use the city field for the borough.



My only criticism of what you said would be "...but users don't see types, 
but words...". Those translations don't have to generate a single word, and 
I would fully expect many to require a short descriptive phrase. As an 
example of this, consider the US which is widely deemed to have 50 states 
(and historically less than that). However, Washington DC isn't in any of 
those States, and is classed as a "federal district" (DC = District of 
Columbia). I have seen people recommend that they just enter "Washington DC" 
as a pseudo-State, and or even a pseudo-County with an empty State, but 
they're technically wrong approaches that are necessary only because the 
associated software was too limited. Having a US-specific type whose 
end-user description would be "federal district" doesn't sound a big problem 
to me.

The new place structure will allow a place to have a type of "federal district".  It will also allow a London borough to have the type of "borough".

Nick.


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

Nick Hall
In reply to this post by ACProctor
On 19/05/14 19:52, Tony Proctor wrote:
> I would prefer to select a region name from a list first. For instance,
> 'England' or 'Netherlands', and then have it present a form with the
> appropriate fields in. If there were some important variation (e.g.
> 'Medieval England' in order to include "Shires") then it could be a separate
> entry on that list. This is how I do it at the moment, and it's nice and
> simple, but some people might want configuration options to control the
> region names that they're presented with, just to make the UI more slick.
>

At the moment the place editor allows the user to define one node at a
time.  They define a name and type.  The title is built by walking the
tree back to the root node.  This works quite well once a few nodes are
defined.  The editor is quite complex allowing multiple parent nodes,
alternative names, source citations, notes and media attachments.

A simple list of fields would be easy for users.  There was a suggestion
a long time ago to provide combo boxes to pick a selection from each
place level.

How would it work for England?

In general, I would want:   England -> county -> city/town/village ->
locality -> street -> building

but I could also have:  England -> London -> borough -> locality ->
street -> building

Would "England" and "England, London" be separate regions?

I'm not sure I can quite see a neat UI yet.


Nick.


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

ACProctor
In reply to this post by Nick Hall
Re: No.  A single list of types was easiest to code.  The user can add to it
by defining custom types.

I think you misunderstand Nick. I still have a single list of types, but
each country only uses a subset of them for its own field designations. As I
said earlier, trying to economise doesn't help genealogists to be accurate.

    Tony Proctor

----- Original Message -----
From: "Nick Hall" <[hidden email]>
To: <[hidden email]>
Sent: Monday, May 19, 2014 9:47 PM
Subject: Re: [Gramps-devel] Place types in v4.1


> On 19/05/14 14:12, Tony Proctor wrote:
>> Ah, that old subject... Thinking "sideways", why does Gramps need a
>> single
>> set of types that applies to every country? Let me explain...
>
> No.  A single list of types was easiest to code.  The user can add to it
> by defining custom types.
>
>
>> Each country has its own divisions/subdivisions, whether geographical or
>> administrative. The list in this thread already excludes some, e.g.
>> Townland
>> (Ireland) and Township (several countries). My approach was to collect a
>> rather large set of such types which included as many countries as I knew
>> of, and then allow a configurable subset to be used for each country.
>> That
>> means I can still use a single common vocabulary, but each country is not
>> forced to utilise every single possibility, e.g. "State" in the UK
>> doesn't
>> mean anything. "Province" exists in both UK (although rarely used),
>> Ireland,
>> and Netherlands but not in the US. However, a State cannot be equated
>> with a
>> Province since they are quite different in concept.
>
> I started with a small list of seven items.  These were the fields in
> the previous version of Gramps.
>
> One option would only have these 7 pre-defined items and let the user
> add custom types when necessary.  I decided to add some extra ones.  My
> test data consisted of English, American and French places.  This is why
> I added Region, Department and Borough.  Users requested Municipality
> and Farm.
>
> I soon realised that the list was getting large.  If I split out
> building types it would get even larger.
>
> Letting the user define subsets of the list for different countries is a
> good idea.
>
> Nick.
>
>
> ------------------------------------------------------------------------------
> "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
> Instantly run your Selenium tests across 300+ browser/OS combos.
> Get unparalleled scalability from the best Selenium testing platform
> available
> Simple to use. Nothing to install. Get started now for free."
> http://p.sf.net/sfu/SauceLabs
> _______________________________________________
> Gramps-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gramps-devel 


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

ACProctor
In reply to this post by Nick Hall
The combo-boxes are essential if you want to retain a consistent set of
place names for each type. This presumes that Gramps distinguishes the
evidential form (i.e. what was written or recorded) from the normalised form
(i.e. the actual name or value). This is something I do for all data,
including person/place names and dates.

Does Gramps impose a strict hierarchical order, or can the end-user put the
name parts in any order they want? I was wondering how it would cope when
there was some administrative change to consider for the time period of
datum. For instance, a town being in one county before a given date, and
then a different county thereafter. This is certainly a problem I've tried
to solve for England as there have been many confusing and unnecessary
changes in the past.

    Tony Proctor

----- Original Message -----
From: "Nick Hall" <[hidden email]>
To: <[hidden email]>
Sent: Monday, May 19, 2014 10:36 PM
Subject: Re: [Gramps-devel] Place types in v4.1


> On 19/05/14 19:52, Tony Proctor wrote:
>> I would prefer to select a region name from a list first. For instance,
>> 'England' or 'Netherlands', and then have it present a form with the
>> appropriate fields in. If there were some important variation (e.g.
>> 'Medieval England' in order to include "Shires") then it could be a
>> separate
>> entry on that list. This is how I do it at the moment, and it's nice and
>> simple, but some people might want configuration options to control the
>> region names that they're presented with, just to make the UI more slick.
>>
>
> At the moment the place editor allows the user to define one node at a
> time.  They define a name and type.  The title is built by walking the
> tree back to the root node.  This works quite well once a few nodes are
> defined.  The editor is quite complex allowing multiple parent nodes,
> alternative names, source citations, notes and media attachments.
>
> A simple list of fields would be easy for users.  There was a suggestion
> a long time ago to provide combo boxes to pick a selection from each
> place level.
>
> How would it work for England?
>
> In general, I would want:   England -> county -> city/town/village ->
> locality -> street -> building
>
> but I could also have:  England -> London -> borough -> locality ->
> street -> building
>
> Would "England" and "England, London" be separate regions?
>
> I'm not sure I can quite see a neat UI yet.
>
>
> Nick.
>
>
> ------------------------------------------------------------------------------
> "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
> Instantly run your Selenium tests across 300+ browser/OS combos.
> Get unparalleled scalability from the best Selenium testing platform
> available
> Simple to use. Nothing to install. Get started now for free."
> http://p.sf.net/sfu/SauceLabs
> _______________________________________________
> Gramps-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gramps-devel 


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

jerome
In reply to this post by jerome
OK, it sounds right.

It seems that I just need to play a little bit with translation strings...

French translation is used by some different countries:
* Canada with a North American culture
* Belgium, Switzerland, Monaco, etc ... and France 

None of these countries are using the same level type or name...

Before 4.1, we displayed something like:

# province (Canada, Belgique)
#: ../gramps/gen/lib/placetype.py:70
#: ../gramps/gui/editors/displaytabs/locationembedlist.py:58
#: ../gramps/gui/filters/sidebar/_placesidebarfilter.py:100
#: ../gramps/gui/views/treemodels/placemodel.py:266
#: ../gramps/plugins/lib/maps/placeselection.py:145
#: ../gramps/plugins/tool/extractcity.py:390
#: ../gramps/plugins/webreport/narrativeweb.py:3366
msgid "State"
msgstr "Région/Province"

# comté (Canada)
#: ../gramps/gen/lib/placetype.py:71
#: ../gramps/gui/editors/displaytabs/locationembedlist.py:57
#: ../gramps/gui/filters/sidebar/_placesidebarfilter.py:99
#: ../gramps/gui/views/treemodels/placemodel.py:266
#: ../gramps/plugins/lib/maps/placeselection.py:146
#: ../gramps/plugins/webreport/narrativeweb.py:373
msgid "County"
msgstr "Départ./Comté"

# province (Canada, Belgique)
# trunk
#: ../gramps/gen/lib/placetype.py:76
msgid "Province"
msgstr "Province"

# trunk
#: ../gramps/gen/lib/placetype.py:77
msgid "Region"
msgstr "Région"

# trunk
#: ../gramps/gen/lib/placetype.py:78
msgid "Department"
msgstr "Département"


So, maybe to generate new translation strings with context (means new msgid)?

Or maybe better, I suppose that 'fr.po' can use 'Province' for 'State' and 'Comté' for 'County'. These levels also existed in France before 1792... :)

I think I can be more accurate on some others levels.

# trunk
#: ../gramps/gen/lib/placetype.py:80
msgid "District"
msgstr "District"

# trunk
#: ../gramps/gen/lib/placetype.py:81
msgid "Borough"
msgstr "Arrondissement"


Districts also exist

I suppose I will also have fun with Canada...


Otherwise, the new place hierachy works fine.
Thanks!


PS: I can also switch between different countries like France & Germany (world wars & co) according to periods. My ancestors' places are grateful!


Jérôme


Le lun. 19 mai 2014 at 19:55, Jerome <[hidden email]> a écrit :
Nick,

About translations, it seems that we are creating custom types!
So, maybe a minor GrampsType handling issue?
We already had such issues in the past. 
I will look at this.

------------------------------------------------------------------------------ "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE Instantly run your Selenium tests across 300+ browser/OS combos. Get unparalleled scalability from the best Selenium testing platform available Simple to use. Nothing to install. Get started now for free." http://p.sf.net/sfu/SauceLabs _______________________________________________ Gramps-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gramps-devel

------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

jerome
In reply to this post by Nick Hall
Nick,

If 'Department' and 'Region' have been only added for France, then maybe we could ignore and remove them from pre-defined items!

Current equivalence (Belgium - Canada - Switzerland - France) for 'County' is working with 'Départ./Comté', where comté also existed in France.

"France
A comté was a territory ruled by a count (comte) in medieval France. In modern France, the rough equivalent of a "county" as used in many English-speaking countries is a department."


Same for 'Region'.

An US State is like an european country, but "Région/Province" fit not bad for most french speaking areas and France.


To have subsets for different countries could be difficult to handle in portuguese, spanish, french, etc ... where the language, definition and wording could be really different according to the country. 

+ 1 for custom types!

Maybe we can remove 'Region' and 'Department' from the default list.

I have an other problem with 'Borough', where the use in Canada seems close to an US district!


Jérôme

One option would only have these 7 pre-defined items and let the user add custom types when necessary. I decided to add some extra ones. My test data consisted of English, American and French places. This is why I added Region, Department and Borough. Users requested Municipality and Farm. I soon realised that the list was getting large. If I split out building types it would get even larger. Letting the user define subsets of the list for different countries is a good idea. Nick. ------------------------------------------------------------------------------ "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE Instantly run your Selenium tests across 300+ browser/OS combos. Get unparalleled scalability from the best Selenium testing platform available Simple to use. Nothing to install. Get started now for free." http://p.sf.net/sfu/SauceLabs _______________________________________________ Gramps-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gramps-devel

------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Place types in v4.1

jerome
In reply to this post by Nick Hall

Just a question: does "Enclosed By" into placerefembedlist means something like "Ruled by"?

I can guess some possible custom types could be 'Duchy', 'Seigneurie' (Europe or Canada) , 'Fief', 'Margrave', 'Principality', etc ... eg, for Germanic culture, where these areas were often used (until half-19th century).


The 4 marked with (*) are in previous versions of Gramps. Again, do we need them all? Anything missing? E.g. Territory. Any comments? What problems does this cause for translators? Regards, Nick. ------------------------------------------------------------------------------ "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE Instantly run your Selenium tests across 300+ browser/OS combos. Get unparalleled scalability from the best Selenium testing platform available Simple to use. Nothing to install. Get started now for free." http://p.sf.net/sfu/SauceLabs _______________________________________________ Gramps-devel mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/gramps-devel

------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scalability from the best Selenium testing platform available
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
12