SPOT Project Java Classes

Download Report

Transcript SPOT Project Java Classes

Final Project Status Presentation
May 1, 2003
SPOT
www.cs.utexas.edu/users/cs378spot
Li Geng, Akbar Noorani, Aimee Ronn,
Stephen Sisk, Mollie Wendel, Billy Wu
Context and Objectives
 The customer’s problem
 Need to keep track of daily practice of XP development team
 Need to generate statistical data to support productivity of
XP practicum
 Need to view the result transparently by various
stakeholders
 Product development goals: An automatic integrated
tool which can:
 Retrieve data from data file (XML)
 Analyze data and generate categorized information
 Provide a front-end interface for developers/users to view
the result
 Specific development objectives
 A scalable and portable product
 Apply XP practice in the product development
Accomplishments/Progress
Customer and Development Team Satisfaction
100
90
80
70
Satisfaction
 It works! (mostly)
 The wave, crazy
Billy dance, Jello
 IDE, talking with
devs, semiavailable
customer, good
teamwork
 Informing the
customer is
important
60
Customer
50
Development Team
40
30
20
10
0
Iteration 1
Iteration 2
Iteration
Iteration 3
Final Statistics and Data
 We delivered 16 stories over the 3 iterations
– We did not have significant scope changes from
our customer
 Total hours worked: 415
 Quality measures - total # of defects: 2.
1 fixed, 1 still open.
 Iteration schedule plan – 1st: behind;
2nd: on time; 3rd: on time
 Product size - # classes: 16, # SLOCs: ~2000,
# system tests: 10, # of unit tests: 27
Issues and Risk Remaining
 Due to the limitations imposed by GNU plot, we were not able to
exclude weekends in our graph even though they were excluded
in our calculations of mean and standard deviation.
→ Should be fixed by using some other graphing software.
 Need to have a Master XML file that we can test all of our code
on.
→ Need someone to debug Mollie’s Test File.
 If the actual work done is more than 10 times greater than the
expected work, then we have an overflow and our Earned Value
Graph has one flawed number which changes the appearance of
the entire graph.
→We believe this to be a GNU plot error.
☺ No outstanding technical Problems.
 Technological Dependencies: GNU Plot.
☼ Don’t see this as a big problem.
Lessons Learned
 Development process lessons: design &
refactor.
 XP lessons: pair-program & test.
 Customer interaction lessons: Elvis was
right.
 IDE tools experiences: intelliJ addiction;
just because it is an IDE doesn’t mean
everything is integrated; when the IDE
fails…
 Final “do differentlies” list: refactor; test
more; plan a test strategy; celebrate!
Future Work
Customer usage and evolution
requirements anticipated
– Package Defects web page
– Unit Test Count per Day web page
Goals for any known follow-on work
– Finalize the XML schema
– Being able to access the rest of the
data fields in the XML file