GeoNames Forum

In a practical use of the Geonames data, a project I am working with has a number of members who are in Turkey.

Several of them have written to complain bitterly about the classification of Turkey under "Asia" instead of "Europe"

Is there a reason why Turkey is classified under Asia and not under Europe? Is this essentially a political argument between Western Europeans and the Turks? Was the decision arbitrary based on geographic proximity, or is there a political justification?

I realize myself that there are always going to be territorial disputes and disagreements in some areas of the world (Taiwan and China come to mind). However in this case I am wondering what the thinking process is, or if Turkey should be reclassified to Europe by some logic that the Turks have?

bernard wrote:

Other suggestions :

Coordinates (0,0)

Mount Everest, North (or South) Pole, or any other extreme point. See http://en.wikipedia.org/wiki/Extreme_points_of_the_world

Also : United Nations Building, either in NY or Geneva

United Nations designation is a bad idea in my opinion. The world itself is first and foremost a geographic entity, not a political entity. If this project were the "Political Map project" then perhaps that could be debated (although it should be noted that the UN charter itself does not disavow the sovereignty of nations and therefore should not be considered a supreme world government of some sort).

I think coordinates 0,0 is a nice idea.

marc wrote:

The difference is that VI is modeled as ADMD (administrative division) whereas VG is a PCLD (dependent political entity).

Maybe it would make sense to model the US dependencies as PCLD too. What do you think?

Best

Marc

Although I cannot claim expertise on the political question, my implementation challenge is trying to get a complete list of child entities within the North American division.

My implementation depends on people clicking through a hierarchical structure where children are retrieved (and cached of course in my database) and then displayed as sub-item options.

If I cannot pull the North American child data to find the U.S. Virgin Islands, how can I reliably locate all items which claim parentage of North America? That seems difficult if not impossible if parent-child relationships are not declared both ways.

Just another comment on this subject: I suppose I understand somewhat the logic of ADMD and PCLD as it concerns toponyms that fall within the territorial boundaries of established political entities. However, in the case of toponyms that might have both a political parent entity as well as a literal geographic parent entity (of some importance), might it make sense to do the following?

1. Establish a single heritage hierarchy for the toponym in question, based on its political affiliation and identity.

2. Establish it within the children data for both a) the political affiliation, and b) the literal geographic domain ("especially" in a case like the U.S. Virgin Islands, where they have an ISO-issued 2-letter country code as well as a separate physical location outside of the U.S. proper).

I understand this is a potentially difficult situation, but my main concerns at the moment is a practical one -- how may I reliably obtain all country-level child entities that reside within the "North America" toponym, whether or not the larger question of multiple inheritances is solved.

I'm not sure what to suggest exactly. Perhaps you are right to treat it as the British case, but since I have no clear idea of the politics of this, I am unable to offer a recommendation myself. Perhaps I could do some research if the situation remains unclear, but for now, I am looking for a practical solution to my immediate problem (if that were possible).

U.S. Virgin Islands (country code VI) shows its parent as North America, but is not a child of North America (if you query North America for its children).

Example:
http://ws.geonames.org/hierarchy?geonameId=4796775

This seems like an omission? Shouldn't U.S. Virgin Islands be a child of North America (just as British Virgin Islands is)?

marc wrote:

It is a tricky question.

As I understand it 'New York City' is something that encompasses the adm2 'New York County' etc, but it is below the adm1 (state) 'New York'.

The rather primitive feature code system is probably not powerful enough to reflect complex situations like this one. We cannot declare 'New York City' to be adm2 because it is not, it is not a county.

A similar problem exists for other metropolitan areas like 'Greater London'.

I don't know yet how so solve the problem and model the situation satisfactorily. It would be easy to come up with an all purpose data model that is able to deal with all kind of situations. This is not the point. The real trick is to keep it simple and easy to use.

Marc

On thing that maybe can be solved? -- The location "Manhattan" seems completely lost in the hierarchy. At least for now, would it not make sense to make "Manhattan" an entity underneath "New York County" or even "New York" (the city)?

Not certain, just wondering...

Thank you very much, I received it.

It's strange, I seem not to get any emails at all from geonames.org -- it is the only domain I have that problem with. Definitely not in my spam folders, so I don't know how to explain it. If you get a bounce back maybe we could trace the headers... or if you can do a delivery attempt from your server maybe it will give a reason?

Anyway, I guess I will just have to keep the same password for now. :o)

Regardless of any explanation (I assure you, it is not in the spam folder) may we please pursue a practical solution?

May aim is not to prove or disprove the explanation beyond any doubt -- my goal is to gain access to my account.

If you want me to phone you personally, send you my credit card information, or whatever is necessary, please tell me, so that I can get the password reset.

This is rather important to me, just tell me how I can help get it accomplished.

Thank you.

Hello Marc,

No really, I assure you, I have been trying many many times different ways, checking my spam folder, all the usual suspects, but nothing is working.

I also am still able to go to the /account/ page and it says I am logged in as "marc" (a problem I reported to you one year ago).

I assure you there are problems in this area, and I would be more than happy to help solve them if you wish. I can do that on a volunteer basis or call you directly in Switzerland, whatever you require, but this is rather important to me, so I am quite willing to help in any way necessary.

I assure you it is not working for me.

Help... is anyone home?

I've been trying for the past two weeks to reach someone -- contact information on the site is hard to find -- I am a paid commercial user, and I need to update my account, but I can't retrieve my password.

I have tried over and over to use the "Forgot my password" option and, although it appears to work, I never receive an email or any way to reset the password.

Please help! I want to keep my account up to date, but I can't do it...

I understand that when searching for a name, it sometimes doesn't exist in the GeoNames database, but then an alternate lookup is offered from the Yahoo Geocoder lookup service.

Wouldn't it be helpful to offer logged-in members a one-click way of recommending that item be added to the GeoNames database?

It's wonderful that one can add an item to the GeoNames database, but making this a one-click process (for logged members) could really help.

Why is there a child entity of "REGION ΠΕΡΙΒΟΛΙ" under Europe? Is this correct or a mistake?

Indeed your advice was correct, Marc. I was using the free server instead of the commercial server in one of my code classes. All working great now, thanks!

I am trying to query hierarchy and children but getting a note that the service is temporarily down.

Is there a status page where I can check to see announcements of the nature of the problem and how long it might be, before service is restored?

I know you guys are doing everything you can but this would be nice to know, even more important than many new features from my point of view. Service uptime is important (especially since I am a paying customer).

You are right that the simple name search returns a correct result, but what about the postalCodeSearch here:

http://ws.geonames.org/postalCodeSearch?postalcode=75014&country=FR&lang=en&maxRows=10

http://ws.geonames.org/postalCodeSearch?postalcode=75015&country=FR&lang=en&maxRows=10

Merci,

geonameguy

Paris zip codes of 75014 and 75015 seem to be reversed in the database. If you attempt 75014, the point lands on area 75015. Also the reverse is true.