My poor brain is swirling around in a sea of Google eccentricities. The new map program is focused on the area I live in, Teton Valley Idaho, and I recently came across an amazing dataset of 10,000 lat/lng sets that represent every address in our valley.
I have no idea how the data reached the internet, especially since surrounding areas have none, but I suspect it had to do with an earlier address conversion that occurred here. In any event, I have it.
The data initial consists of city: (Driggs, Victor, Tetonia, Felt and Alta) and the Longitude and Latitude for each address. This was enough for me to go banging the lat/lng’s up against the Google reverse goecoder, to obtain additional information like street address and bounds.
Very quickly I ran up against Googles rate limits. They don’t mind if you write an app that sends a location to their API and get an address back within a web app, but when you start sending thousands of them at a time, they have a problem. Actually, they like apps that maintain their addresses, it saves them overhead, as long as you implement the data on their Google maps. You just have to play their complicated game, to get it.
I’ve created a MySql database on my server that is gathering up all of the geolocation data for the Teton Valley as I write this. I had to go through an amazing number of poorly documented hoops from Google including signing up for their pay as you go plan ($.50 per thousand hits).
All of that is done and I have an amazing database of mapable addresses for Teton Valley at my disposal and a very cool map program is in the wings…
Update (10/14/17): All of this is now gone…