CS 481 - Team 49
Final Presentation Minutes
 
Final Presentation Meeting
Wednesday, December 10th, 2007
10:30 - 11:40 AM
JEB 324
Recorder: Shruti Upadhyaya
--------------------------------
 
Summary: This meeting consisted of the final presentation and discussion of the features and defects present in final release. We also talked about what we learned during the semester and presented suggestions for improvement for next semester.

Attendees
     Brandon Riggers, Peter Brown-Hayes, Travis Thoroman, Shruti Upadhyaya, Bill Junk, Bruce Mayes (over the phone).


Topics discussed:


Vision Statement
- Peter talked about how difficult it is to meet the organization’s standards when we do not understand everything.
- We learned how important planning is. During this semester, we made a few changes in the project and planned accordingly. 
- Bruce talked about the subtle requirements that are not documented.
- Bill said he would like to have a performance evaluation data to see where time is being eaten up. 



Next Semester/ Suggestions for Improvement
- Bruce said that the key part of next semester’s project is going to be formatting filelist_turmoil for better results.
- Threading processes as much as possible instead of using linear style. 
- Look for other ways for checkout process to make it faster.
- We weren’t able to give the sponsor as many updates. Giving more updates so everyone knows what is going on. 
- Bruce asked if there are better ways to do investigation in future as lot of time was spent in investigation in the beginning of the semester.
- One of the problems we had in the beginning of the semester was that no one knew where we were heading. It was also difficult to understand the older tool.
- Best way to start in future would be to test a few more things as we are researching different tools. 
- Good documentation should make things easier for next semester. It shouldn’t take more than a few days to figure out what our tool does and know what direction next semester team need to go.
- Next semester team should work with Subversion first. If they do not spend enough time with Subversion, they will be confused for a long time. 
- Use the common platform as HP.
- It is good to keep track of decisions made during the semester.
- Next semester team will also have to have timing data to find out where the major bottlenecks are.


Risk Analysis
- The test we wrote worked well initially. But we found bugs as we made it bigger.
- We need to plan out the code before we test. 

Defects
- Bruce said that a typical S curve is more normal for defects graph. Less defects at first and more as the project gets bigger.
- The line count of code was 724 with SCS. A graph with defects per line of code (724/7(# of defects)) should be shown to see if the number was reasonable. 

Modular codes
- Having well-defined interface would have been useful in the script.
- We should make sure that the documentation is good. Peter commented that the code comments need to be updated as we go. 

Team Effort
- Bruce said that he was not surprised that we spent 26% of our time in Project Management.




Retrospective
- We learned a lot about project management.
- One of the questions raised was how to get hold of someone in the case of urgency when one’s phone and email are not working?
- Bruce commented that it is more important to have more accurate result than faster result. 

Final Action Items:
- We need to record the “log output file” as “deferred.” 
- Minute needs to be sent out.