From: Otis Bricker Date: 2001-12-16T05:34:31+09:00 Subject: [ruby-talk:28644] Re: John Roth dolt ( Re: A challenge to proponents of Unit Testing. ) Ron Jeffries wrote in news:375DBFADF9B0B92A.4C3EC3A30229C828.76845510EA448689@lp.airnews.net: > On Sat, 15 Dec 2001 13:37 +0000 (GMT Standard Time), > brangdon@cix.co.uk (Dave Harris) wrote: > >>> Coverage is interesting, and some XP teams are using it. >> >>Do you have any figures on the kinds of coverage that result from normal >>XP? Is it 95% or more like 45%? Does it tend to drop as the project ages? >> >>I am asking as a way of comparing XP with other methodologies. > > I have seen reports on the extremeprogramming yahoo list of coverage > approximating 100%. Again, coverage analysis is not an XP core > practice, but people are reporting surprising high coverage for an > approach that doesn't push it. What figures are available for other > methodologies? I'll appreciate any references ... > > Thanks, > > Ronald E Jeffries > http://www.XProgramming.com > http://www.objectmentor.com > I'm surprised that you're surpised. I would have thought that you would almost expect 100% coverage. Refactoring would get rid of unused code. And Unit tests should exercise everything that could fail. Including having allowable failures tested for. If a method given a class of faulty input should throw a JunkDataException, then it would have a test to make sure that it does. While XP doesn't stress coverage by using the word, it seems that its methods stress it to the extreme. Oops sorry about that choice of word but I can't think of a better one. Otis Bricker