Stepwise Integration Plan

One team, G3, will be the first team to produce the integrated version v0.1. This integrated version v0.1 is essentially what they have implemented so far, with a skeleton main() program for further integration. The main() program looks like this:
main()
 {  ask the user whether he wants to accept input
    directly from the Kinect device, or read input
    from a pre-stored Kinect file.

    If the user wants to read input from a pre-stored
    Kinect file, ask him to enter the URL
    /*Here we really should be using SIS to send message
      to Kinect server, which will return URL to us.
      But lets do it the easy way "/

    ask the user to select an option from the eight
    groups g1 through g8.  This is a switch to invoke
    one of the eight programs
  }
In v0.1, team G3 will of course only have g3() to test, and the other seven are just stubs. Once they get v0.1 working, the other groups can take v0.1 and expand it to produce other integrated versions. I will organize the groups into a logical sequence, so that the final v0.8 will become the truly integrated version. If any group g(i+1) fails, the next group g(i+2) can start version version v0.i and side-step the non-working version to produce version v0.(i+2).

Thus once G3 succeeds, G3 can put all source codes of v0.1 in one place at their website, for next group to take over, and G3 is done.

The other groups will follow the same procedure, until all are done.


Group 3 completed the main skeleton for the class project. Instead of using the file url Group 3 decided to use the actual SIS component.

The main component is implemented in Java, so groups that are using Java can just call their components directly, and groups that use C++ can implement simple JNI stubs. Group 3 included a simple example (grouptest) for groups using C++, and since their component uses C++, it can also be used as an example.

The source code can be found online here: http://code.google.com/p/kinect/source/browse/#svn%2Ftrunk%2Fproject.