To get going:

git clone git://github.com/danvk/sfhistory.git
cd sfhistory
dev_appserver.py viewer

Then hit up http://localhost:8080/


If you want to do your own geocoding, you'll want to do this.
gunzip -c records.pickle.gz > records.pickle
tar -xzf geocache.tgz
echo "records.pickle\ngeocache\n*.pyc\nimages" >> .git/info/exclude

These files could be checked in directly, but that seemed wasteful.

Python can read the pickle.gz file directly, but this winds up being an order
of magnitude slower than reading the uncompressed version.


To start up a new local AppEngine instance:
- (open up the GoogleAppEngineLauncher and start up the app)
- Run "./upload_to_appengine.py" and Ctrl-C when you're ready
- Hit up http://localhost:8080/


Here's the sequence to bring up an instance of the viewer app:

# Geocode all images that can be geocoded:
$ ./generate-geocodes.py --geocode --maps_key (your key) > /tmp/geocodes.txt

# Convert these geocodes into JSON:
$ ./generate-js.py > viewer/lat-lons.js

# Upload JSON to AppEngine
$ cat /tmp/password | appcfg.py update --passin --email danvdk@gmail.com viewer

# Upload (new) thumbnail images to AppEngine
$ ./upload-thumbnails.py