+1 for removing no-controlled System.*.println. I believe there is a precedence of using a "Test" log channel in existing tests that can be used to log debug type info.
Albert Lee. On Wed, May 27, 2009 at 10:26 PM, Craig L Russell <[email protected]>wrote: > > On May 27, 2009, at 8:02 PM, Michael Dick wrote: > > Hi all, >> >> In the most recent commit for OPENJPA-1015 (code conventions for >> testcases) >> I noticed there are a lot of System.out.println statements in our >> testcases. >> >> >> Are the printlns really useful to anyone? The only reason I can see for >> having them is if the testcase requires some manual interpretation to >> ensure >> it passes or fails (generally a sign that you need to rewrite the >> testcase). >> At best they take up time formatting Strings and writing out to the >> console >> all of which is ignored by the build (or builder). >> > > I agree. Thousands of lines of output serve only to hide the real results > of the tests. I'd like to see all System.out.println removed from the test > cases. Replace them with logger.info directed to a file that someone can > look at if they want to see what happened. But most developers just want to > see tests run and mvn tasks succeed. > > Craig > >> >> >> Is there a good and compelling reason why we keep adding these in? Adding >> assert() and fail() checks makes sense - but dumping a bunch of diag info >> by >> default seems wasteful to me. >> >> -mike >> > > Craig L Russell > Architect, Sun Java Enterprise System http://db.apache.org/jdo > 408 276-5638 mailto:[email protected] > P.S. A good JDO? O, Gasp! > > -- Albert Lee.
