Re: [udig-devel] Re: CRS - "Bursa wolf parameters required."

[Martin Desruisseaux <martin.desruisseaux@xxxxxxxxxxxxx>]

Jesse Eichar a écrit :
> How is the database intended to be used?  In the case of udig would
> there be one running instance somewhere in the internet that all udig
> instances hit?  Or does one get started each time udig runs?  

This is at yours choice.

If you want one running instance somewhere in the internet that all udig 
instances hit, you can do the following:

  - Setup a PostgreSQL (or some other) database on some uDig server.
    For the following, I assume that a PostgreSQL server is running
    on some distant server with SELECT permission grant for all uDig
    users.

  - Import the EPSG database in this PostgreSQL server (just run the
    SQL scripts provided at http://www.epsg.org).

  - Cut and paste the code in the box at the following javadoc page
    (section "Any platform installation"). Put this code in whatever
    class and package you want.

http://modules.geotools.org/main/apidocs/org/geotools/referencing/factory/epsg/package-summary.html

  - Edit the server name, database name, maybe add a user name and
    password. Bundle this class in yours uDig JAR (if dependency
    to PostgreSQL JDBC driver is acceptable).

  - Add a 
META-INF/services/org.geotools.referencing.factory.epsg.DataSource
    file (to be bundled in the uDig JAR). Declare in this file the
    fully qualified name of the above-cited class.

Thats all. You can declare an arbitrary number of EPSG DataSources with 
different priority numbers (for example a main server and mirrors used 
as fallback).


	Martin.