Post-Mortem on This Term's Projects
We held an impromptu post-mortem on some of this term’s CSC49X projects on Thursday. The course is a lot more productive than it used to be, thanks in no small measure to feedback from students about what works and what doesn’t. Here are the highlights from three of this round’s teams.
The Good
- On-Line Marking
- Team not discouraged by being behind schedule, and even better, catching up on the schedule
- Good scheduling
- Fun experience
- “We finally clued in” entering details in tickets
- Having something to demo
- Kids Help Phone
- Good teamwork
- Recommend to attend KHP research meetings to find out the other side of the project
- Good to have rollover and Andrew’s previous experience in the project
- Good use of bug tracking and ticketing system
- Proactive team
- UML Debugging
- Learned a lot about relevant Eclipse technology
The Bad
- On-Line Marking
- Communication failures; intermitent, little communication
- Breaking other people’s work when modifying
- Lack of testing because it was left until the end
- Lack of face to face communication: email wasn’t being enough
- Had to deal with sloppy code
- Not enough carry over from previous team
- Final report weighs too much (50%)
- Greg’s response: used to be 100%, and you have to learn how to present your work (especially if you’re going to grad school or starting a company)
- Kids Help Phone
- Interface testing broken by some changes
- Interface testing changed too many times
- Bad titles for tickets and no glossary, which led to difficulty in searching tickets
- No intergroup communication
- Few and large commits, with poor comments
- Didn’t use milestones in time
- Hard to find tickets because we didn’t follow some labeling conventions
- Estimates too optimistic
- UML Debugging
- Didn’t do A&E so it was difficult to track progress and next steps, even for a one-person project
- Course should be more publicized
- Greg’s response: did everything I could, but still having a hard time reaching good students