Update: I build a complete project with Maven now that you can check
out at my github repository. Feel free to
fork the repository and build on it.
Yesterday someone asked me to help out with the first steps with
the Fitbit API and fitbit4j. And although I think
fitbit4j is bloated almost to unusability I wanted to verify my
sentiments and try to build a simple example application that reads out
basic data. This should be a very easy tutorial for the Fitbit API.
Here is what I got after revisiting the help I got from the development
group when I started:
Make sure your fitbit application is setup as a desktop application to
give you the PIN. The output of the application will show your display
name and the registration date on fitbit.com.
This is the most basic application that has access to the Fitbit
API using their own library. Isn’t that a little too much just to
get data from a REST API? They’re forcing a
FitbitApiSubscriptionStorage and the
instantiation of a
my throat. Compare that to the same actions in Ruby:
Much cleaner. Just avoiding all the boilerplate about the caches and
storages would make fitbit4j so much easier to use. Well, that is
exactly the reason why I decided to lern myself some Ruby On Rails
while building the fitbit analyzr. No java webapp, no fitbit4j.
Since then I never looked back and I’m happily building my application
with Rails. 🙂