Ben Reichelt’s Weblog

We have a client application at work that we give to our
clients so that they can sync their contacts and appointments from our system
into Outlook.  There is an Outlook api to do this, so its fairly
straightforward. The way our stuff works is the user installs the client
application and clicks the ‘Sync’ button.  The client app then accesses a
web service that we have exposed that will return the users contacts and
appointments in the form of datasets. 

The client app then processess the
app and add the necessary information to Outlook.

It would seem, based
on Rockford Lhotka’s definition of a web
service , that we are
misusing webservices.  I know that web services are most useful at interop
between different systems, but I think they can also be useful, even when you
are targeting specific uses of your web service. Using a web service could just
happen to be the easiest way to transfer information, as in our case. 

The
client application COULD make a direct connection to the database, but using a
web service as an intermediary seems a little more secure, IMO.

We could also write an entire client/server application that
transfers the data in an application specific way, but that would take a
considerable amount of time compared to the ease of making a simple web
service.

So, while other applications might have a difficult time
using our exposed web service, that was never the intent, and I think thats an
acceptable reason sometimes.