Greetings, all,

I woke up this morning to the following in my mailbox:

------------ Forwarded Message ------------
Date: Friday, April 15, 2005 2:00 AM -1000
Subject: Hackystat Daily Report: 2005-04-15

The following alert(s) were triggered today:
* Project Daily Summaries
hacky2004-all on 14-Apr-2005
 Active Time: 8.75 (total), 0 (yours)
 Build: 65 (total), 60 (successful)
 Coverage: 78.4% (overall), 3634 (covered methods), 1001 (uncovered
methods)
 File Metrics: 1328 (classes), 6333 (methods), 91643 (LOC)
 Unit Tests: 2245/5/1 (total p/f/e), 0/0/0 (yours)
 Churn: 1905/876 (total lines added/deleted), 0/0 (yours)
 Commits: 70 (total files committed), 0 (yours)
 Issues: 45/0/28/132/207 (o/i/r/c/t), 25/0/1/27/54 (yours)
 Performance: 52 (total tests), 8 (failures)

---------- End Forwarded Message ----------

I always skim through this report, and noticed two strange things about
this one:

- LOC has jumped over 90K.  It was in the mid-80s yesterday.
- Coverage has dropped below 80%. It's been slightly above 80% for years
now.

So what I did was bring up two daily summary reports, one for the 14th and
one for the 13th and compared them side-by-side.  What I learned was that
hackyPRI started appearing in hacky2004-all yesterday, which added 5600
lines of code to this project. That's fine.  What's not so good is that it
added added 223 uncovered methods (the highest number of uncovered methods
of any of our modules!)  That's what accounts for the sudden significant
drop in coverage.

Aaron, can you take a look at the situation and see if you can improve the
testing on this module?

Thanks,
Philip

p.s. This is a _classic_ example of how having telemetry data can support
project management. Yeah for us!

Reply via email to