The Geographic Extensions for TurboGears makes is easy to add GIS capabilities in a TurboGears2 application. A Web GIS application typically consists of both server side and client side components.
The client side component typically consists of a Web2.0 based map with rich Ajax capabilities for panning, zooming, layer selection, etc. The tw.openlayers ToscaWidget library makes it really simple to add a map on a TG2 application. However, if the client side application becomes complex and involves a lot of javascript work, it is advisable to use the the OpenLayers library directly for greater flexibility.
On the server side a GIS application needs to handle HTTP requests for query, processing and manipulation of GIS objects. Several server side GIS tools already exist in the python world and tgext.geo makes it easy to integrate these tools for server side processing. The tools integrated by tgext.geo are:
GeoAlchemy is a SQLAlchemy extension for spatial databases. Use this to define you model objects which have geometry fields. The FeatureServer Tutorial explains the usage. For further details refer to the GeoAlchemy docs.
FeatureServer is a GIS server that supports publishing of GIS data from multiple datasources including GeoAlchemy. FeatureServer can publish GIS data in several formats such as JSON, GML, RSS, etc.
MapFish is a pylons based GIS server. MapFish also has a rich client library that uses the mapfish protocol. So, if you want a complete WebGIS solution, MapFish is for you. MapFish model and controller definitions can be added to TG2 application using the paster commands. As of now MapFIsh model definitions do not use GeoAlchemy and support only reflected tables. However, work is on for building future versions of MapFish on GeoAlchemy.
Serving raster data (satellite imagery, etc) is not supported by TG2 as this is not a very common use case. Most of the time we are just interested in requesting them from public / corporate servers and displaying as background layers on our maps. However, in doing so we could choose to cache the requested tiles (view elements) on our servers (or in the cloud) in order to make our application more responsive.
TileCache is an excellent python library for achieving this. tgext.geo provides a paster command to integrate TileCache in an existing TG2 app. The details are available on the TileCache home page and on the TG2 TileCache Tutorial.