Phlip wrote: > What I meant was TDD requires frequent edits and test runs, and the > handle exhaustion problem is causing frequent reboots between them! I > don't understand how anyone else can use Watir for anything, much less > TDD. > I run Watir tests all the time. And i actually often use TDD to drive the development of both Watir and the testing framework i use in my day job. Mostly i reuse browsers rather than close and open them. And in fact, i go to a lot of trouble to recycle browsers rather than throw them away. We have often had reports of problems from people who have tried to have each test create and then close its own browser. After running tests many many times, i do find that IE starts behaving badly and i need to reboot.
Recently i actually had to have the tests in a particular suite each open and then close its own browsers, and for this i wrote the new IE.new_process constructor that starts a new process for each IE window (you should use IE#kill to close these windows). I believe that this approach avoids the problem that we've seen with the traditional IE.new/IE#close approach. It might also avoid the problem you are seeing. > Maybe everyone has more handles than me. > Can you tell me how i would be able to tell that Watir was having handle problems? How would i measure how many handles are being used so that i could tell whether a particular change fixed this problem or not? Earlier you said that you had a "wait" problem and were thus using Watir_Simple because it managed it for you. Watir_Simple was written for Watir 1.0 and included some additional calls to wait that were missing in the core Watir software back then. But i have been under the strong impression that these are all redundant with Watir 1.4. Bret _______________________________________________ Wtr-general mailing list Wtr-general@rubyforge.org http://rubyforge.org/mailman/listinfo/wtr-general
