I wanted to work on the data-querying first, so we needed to [load the GNIS data].
The GNIS data is in PostGIS and the table looks like this:
Table "public.gnis" Column | Type | Modifiers ---------------+--------------+----------- full_name | text | full_name_lc | text | sub_national | text | country_name | text | country_code | character(2) | est_pop | integer | type | text | uniq_featcode | integer | the_geom | geometry | Indexes: "gnis_spatial_idx" gist (the_geom) "name_indx" btree (full_name_lc)
Note that the data is indexed on the full_name_lc column and there is a spatial index on the geometry. These are crutial for performance.
Here is an example row from the table:
full_name | full_name_lc | sub_national | country_name | country_code | est_pop | type | uniq_featcode | the_geom ----------+--------------+-----------------------+---------------+--------------+---------+-----------------+---------------+-------------------------------------------- Vancouver | vancouver | British Columbia | Canada | CA | | Populated Place | -575268 | 0101000000F0BD648888C85EC00000000000A04840
We have to configure GeoServer to find this data.
I won't repeat how to set up GeoServer here, but I will point you to the tutorials.
We call the layer 'gnis', and the namespace is 'topp' (The Open Planning project). The namespace is just a label before the data, usually signifying who is supplying the data. You can call it whatever you want, even "brent_rules".
>>Proceed onto the next step, Step 3: Set Up the Queries* >>