1.5.1. Extended Client-Server Architecture
Thick and Thin Clients
The WebPark platform was designed around
the so-called "Client-Server" architecture. The idea of this is that a client,
which might be a web browser on a mobile device or personal computer, accesses
data over the internet from a server which acts as a repositiory for the
information. The client and server communicate using agreed standards such as
those defined by the OGC. Refer to the lesson
OGC Standards and WMS
for more
information about this. Clients can have different "thicknesses" depending on
how much processing work they need to do to present the data provided by the
server. A thin client might request data that is returned as a graphical image
file (e.g. png or jpeg). In this case it needs to do little work except display
the image to the user. Instead the server will have performed all the processing
necessary to draw the data correctly. A fat client will access raw data and need
to perform much of the rendering operations itself. The problem for
location-based services such as WebPark is that they cannot rely on a continuous
internet connection. This makes it difficult to employ a straight forward client
server architecture. WebPark uses the caching technique to dissove the problem.
To decide which data to display the WebPark
client needed to be little fatter than in an always online scenario, but not so
fat that it needed to perform large amounts of processing to render the data.
The illustration below describes the basic pattern, together with some of the
processing operations performed on each side.
Extended Client-Server Architecture used in WebPark