From: Mike <m...@pe...> - 2004-09-28 04:49:54
|
On Mon, 27 Sep 2004, Kevin Goess wrote: > Rita, every time we add a feature or correct a bug in the log4perl > libraries, we're going to add test cases for the new feature or the > bugfix. So, yes, if you want the tests to work on the current > libraries, you need to use the current tests. Ritu, let me add to Kevin's comments: If you write your own appender, you should bundle it in your own distribution and provide test cases for it, independently of Log::Log4perl. Our API should always be backwards compatible, unless for a good reason (this has never happened as far as I remember, and should it one day, only for a very good reason and it's gonna be announced boldly) or, of course if someone introduced a bug :). As Kevin has pointed out, our test suite is subject to constant change, relying on it won't work. Your test cases should be independent of them. Hope that helps! -- -- Mike Mike Schilli m...@pe... |