Devcathlon: Starter Events

This week we are starting to code for the Devcathlon system. What is the Devcathlon system? My pervious blogs will tell you what it is. Soooo… this week, each of us in my software engineering class get to work on a starter event (http://code.google.com/p/hackystat-ui-devcathlon/wiki/StarterEvents).

Starter events are simple events to code to get started on Event portion of Devcathlon. My starter event was:

Pass The Build
Summary: Give points for passing builds.
PAR: Wakeup once per day. Determine how many builds occurred during that interval and how many passed and how many failed. Award 1 point for each passing build during the past day and deduct -1 point for each failing build during the past day.
Decay/Incentives: Award 10 bonus points for 7 straight days with at least one build per day and no failures.

So this event only checks one team at a time and rewards +1 for each successful build deducts 1 for each failed build. It also gives a bonus for 7 straight days of successful builds and no failure. The problem I ran into was going though the correct Hackystat Javadocs for the right line code to work with, I asked some of the classmates for help, which was great. Another problem was the TestConfigurationManager file, it tests the total number of events, and the problem was incorrect number of events to test.

assertEquals("Test events", 7, config.getEventConfigurations().getEventConfiguration().size());

Merge conflicts happened this morning when I was uploading my java files onto main repository. Hudson failed to compile and build because of incorrect number of events and other things. Like around 12:30 AM 2/25/09, Hudson failed for couple of hours, the person that failed it did not fix it. One of the classmates that were waiting for the Hudson to compile and build successful gave up the wait and went in and fix it. One other problem is that coverage tool Emma (http://emma.sourceforge.net/) is showing 0% line coverage for all test files, I do not know what is going on, for example my test cases is working in Eclipse, but showing failed in Emma.


On the side note, I think the screencasts (http://code.google.com/p/hackystat/wiki/Screencasts) are great, if it is possible to get the file too that was with each screencast will be awesome as well.