----- Original Message ----- From: "David Graham" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Wednesday, February 12, 2003 3:32 PM Subject: Re: New Tests
> > >Not exactly, 'running' the tests is the only thing that will take longer. > >Building will actually be quicker with approach #2, because there's less > >code to compile. > > > >Also, since I'm writing these from scratch (not changing anything that > >exists), either technique will increase test time. > > > > Don't the tests run when you build the distribution? No, you have to specifically run the tests (assuming you've setup the proper container config....tomcat.home.33, tomcat.home.40, etc) > > > > > 2. The jsp test is now dependent on other Struts tags' behavior. What > >if > > > one those tags is broken? The goal of testing is to isolate the most > >basic > > > usage which your first example does. > > > >I thought that was the purpose of tests. If doing it this way, we can > >identify bugs that might not have been caught before, that should be a good > >thing. > > Ideally, tests shouldn't rely on other components that could be broken > because then you don't know why the test failed. I don't understand what you mean by other components that could be broken. The only way a test might falsely appear to pass is if <logic:equal> is broken. If there is enough concern, then I can certainly test with a scriptlet: <% if (request.getAttribute("runTest") == "testSomeTag"){%> Thoughts? > Dave > -- James Mitchell --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]