On the one hand, testing against Workbench allows you to isolate Tapestry from RMI and database issues.
On the other hand, testing against Vlib (or this new Pet Store) is a more realistic test and may identify hotspots more accurately. Meanwhile, I'm working to create JUnit tests for Tapestry. I want some real tests in place before I tear it appart for the JSP interoperation. I'm building mock- objects versions of the Servlet API. -- [EMAIL PROTECTED] http://tapestry.sf.net > > Howard, > > > At some point, I need to purchase a good profiling tool > > and see what's really going on inside Tapestry. There > > I have OptimizeIt 4.x and can offer to do some profiling if we can isolate > repeatable tests. I could run the unit tests several times (to accomodate > for HotSpot), but since that's not a typical setup compared to an > application it might not be too useful. Alternatively I could stomp on the > Workbench with HTTPUnit, or even a real-world project for which we (more > precisely: Christian Noack) have HTTPUnit tests. > > > does not, yet, appear to be a useable, free profiling > > tool. > > No, unfortunately I don't know any, either. > > > Of course, CPU power is the one thing that you can > > always buy more of, and it gets cheaper every day! > > I'm afraid I have to disagree with that. If you can spend $0 on faster/new > hardware, it doesn't matter how cheap it is. Personnel and installation > costs for the new machine are sometimes more, if not the deciding factors; > not to mention getting the purchase order through management. > > regards > Holger ------------------------------------------------------- This SF.NET email is sponsored by: AMD - Your access to the experts on Hammer Technology! Open Source & Linux Developers, register now for the AMD Developer Symposium. Code: EX8664 http://www.developwithamd.com/developerlab _______________________________________________ Tapestry-developer mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/tapestry-developer
