Fix TreeModel.print() method implementation

[bhufmann]

The TreeModel.print() was implemented initially as prototype and optimized for DATA_TREE types and in particular for the function duration statistics. It uses the ColumnDescriptor field that had been provided by the server, but it's not serialized anymore because the TSP doesn't specify it. Instead there is a header field with 'name' and 'tooltip' (see also #33).

This task is track the following things:

1. Track the update so that only data is printed that is provided by the TSP. This means to ignore parameters that are serialized into in 'others' of each object.
2. Since each entry in the TreeModel also has an unique ID to be used to identify the entry and which can be used for further queries (e.g. fetch series data of an xy chart or states for timegraphs). The print method shall print the Id in a separate column.

Note, that server updates are needed so that each data provider complies with the TSP and only TSP-specified fields are serialized.