Peter Chapman finished up some remaining functions in the canopy_stratum class. He also implemented classes that will contain the 'default' objects, and converted the private member functions of the patch class. The entire patch class (most likely) will be converted by the beginning of the week.
Maria Doyle finished converting functions of the canopy_stratum class. She also worked in converting private member functions of the patch class.
Hank Hoffmann finished the header files on schedule due to an early start. He has written 17 header files totaling approximately 2000 lines (including comments). In addition to the header files, Hank implemented the layer class in the new model since this new layer class has functionality that was non-existent in the old model. Hank also sent the client a list of the latest discrepancies between our output and the expected output. Finally, he posted the completed header files and the picture of the object hierarchy on the web for the client to view.
Jack finished working on building the command object from the GUI. This includes the implementation of a command parser. The command parser is used to parse the command line and make it into a command object, if the user choose to manually input the command line from the the GUI.
Skip continued to work on the Java Server, exploring various possibilites to implement the communication between the GUI, server, and model. He also talked with Dr. John Smith about the requirements of using Java Native Methods.
The greatest accomplishment this week was the timely completion of the header files. So far the C/C++ code conversion is on schedule. However, we have yet to figure out how to test/integrate the patch class into the existing code.
At this point in time, Skip believes that the secondary requirement of visualization is not going to be met. The team does not have a basic grasp at this time on the design for accompolishing visualization, given the state of the model as was given to us.
Also, the Java sub team maybe slightly behind schedule in terms of designing the architecture that allows the user to create the necessary input files to run with the model. Also, the team is unsure about certain concurrency issues that may arise in regards to multiple users running model at the same time.
The issue of how we integrate the Java code with C++ code, and test the system as a whole once both parts are finished remains an issue unresolved at this time.