| Author |
Message |
|
|
Sam
The GeoNames dump files are all using UTF8 encoding.
Marc
|
 |
|
|
Laurent
I found the problem with the java api. It is a bug in the java client. I will release a new version in the next days. In the mean time you can grab the csv head as the bug is already fixed there.
The database is available in the form of a postgres dump as part of the commercial offering (somehow we have to make a living) :
http://www.geonames.org/products/reverse-geocoding.html
Marc
|
 |
|
|
Hi Jacopo
We could boundary vector information and do the update automatically. Unfortunately I don't know of a source for good boundary polyons for Polish administrative divisions. Do you know a source?
I have fixed the duplicate ADM1 entries. This should be ok now.
Marc
|
 |
|
|
Yes, you are right they are out of sync. The file in the dump has more countries and the file in the csv has more columns.
If you create a merged version then I will use it also for the dump file and we get rid of the inconsistencies once and for all.
Marc
|
 |
|
|
Hi Laurent
These are two different databases. We have not yet had the time to connect them with their respective ids.
I don't understand the second question. Which service are you referring to with 'the web service API'?
Marc
|
 |
|
|
you are right. I will change it.
Marc
|
 |
|
|
Thanks Oscar.
This was a bug with the minus character treated differently in the search and the name_equals search.
It has been fixed.
Marc
|
 |
|
|
Paulo
I have now also updated the 29 codes (from 10) after making sure there are no 31 among the former 10 codes.
Marc
|
 |
|
|
The '00' code means the code is not known. You could try to help us figure out whether the subidivision does exist, whether it is a ADM1 (or something else?) and which is the correct code.
Marc
|
 |
|
|
cool.
I have added gpuViewer to the users list : http://www.geonames.org/users.html
|
 |
|
|
Shahzad
most files are more or less static or redundant. The files allCountries.zip and alternateNames.zip are updated daily.
Marc
|
 |
|
|
|
Die Datei wurde neu exportiert.
|
 |
|
|
Jhonathan
The alternate names are in the file alternateNames.txt
Cheers
Marc
|
 |
|
|
you have an iso language code in the table alternate names. Join the two tables over the geonameid.
Marc
|
 |
|
|
Hi Shahzad
The message queue is not (yet) available to third parties. I am scared of having to provide support for it and the daily dump is as good or better for nearly all applications.
The timezone info should be correct for nearly all places. I cannot exclude that for some places where strange rules apply the timezone might be wrong for an hour or part of it.
Marc
|
 |
|
|
Hi Shahzad
The data is updated continously :
http://www.geonames.org/recent-changes.html
You can grab a new dump periodically. We have implemented the message queue to synchronize our own servers.
The relationship between cities and countries is in the column 'country code'.
Marc
|
 |
|
|
This is different issue. The granularity for Portugal of our geocoding is not as detailed as the number of postal codes. A lot of postal codes share the same coordinate estimation. The code you are using for the findnearby is just not included in the first rows returned of codes which share the same lat/lng estimate.
I cannot comment on the other countries as I have not enough information on the type of issue you see.
Marc
|
 |
|
|
geonames.de is another project. It has nothing to do with geonames.org
You can help us add Arabic names for places (click on alternate name) :
http://www.geonames.org/2487134/mostaganem.html
and it will turn up in the web service.
Marc
|
 |
|
|
There is no webservice for translation of language names. First it is probably included in most modern programming languages (like java) and it has not much to do with names of places.
Marc
|
 |
|
|
The findNearby service is not meant to return addresses. It is meant for higher granularity like city names or neighbourhoods.
Marc
|
 |
|
|