We could think of a preferred way of how to configure the security model and run the tests for that configuration as part of the drt. The other tests could go into a separate test directory and be run on request, e.g. if someone would like to use a different security model. Comments, suggestions?
Starting Tomcat once during the drts shouldn't be a problem. I'm working on adding that to the drts anyway because we need some end-to-end tests for the WSM project, i.e. we need to run a WS and then test a client against it, to make sure that everything really behaves the way it should. Thus, we could launch it once and run all runtime tests at once. Cheers, -michael -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Monday, December 06, 2004 10:57 AM To: Beehive Developers Subject: Re: @SecurityRole and @SecurityIdentity Hi Michael, That's good to hear providing a "Beehive way" to restrict access to Web services. I've just started to work on the test cases, but I'm thinking how to make the test cases for runtime. To switch the security model, server-config.wsdd should be modified so that it means it requires tomcat be restarted three time since there're currently three security models. Furthermore, the client code should be generated and compiled dynamically three times as well. I don't think those tests involved in drt is smart way... Any better solutions ? Thanks in advance. Wolfgang
