Hi Dave, You're the first one to report this issue. Congrats! :-)
It seems to me the only case where we have a real problem is when performing server-side XSLT. If the returned data is XML or if the XSLT transformation is done in the browser then we can easily implement your solution 1 (i.e. send back data as soon as a test finishes). Would that be good enough for you? I'd rather do simple things than start playing with threads and committed/uncommitted responses tricks... :-) If you agree on this, the next step would to provide a patch for it :-) Thanks -Vincent > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > Sent: 03 February 2004 05:31 > To: [EMAIL PROTECTED] > Subject: Browser/Web Server timeout when accessing ServletTestRunner for > long tests > > > > We use cactus to test our internal application and it is called as part of > our > build script which works well. We have also made use of the > ServletTestRunner to > access the test cases through a custom JSP on our applications website. > This > allows us to run our tests from any browser at any time. However, when we > access > a large test suite, it times out ("The page cannot be displayed") because > it > takes a number of minutes to run all the tests. We could modify timeout > settings > on both the server and the client (I believe in IE it is a reg edit) but I > would > like to be able to call the tests from any browser so need a server > solution. > > Has any one else come across this issue before? > > I can think of two solutions: > > 1 Send information to the browser while the tests are being run which > would stop > the timeout occurring. > By modifying the Servlet Test Runner I could add another TestListener that > writes to the response as each test completes. > > 2 Modify Servlet Test Runner (or create a new runner) that can run in a > asynchronous mode. > This would involve spawning a thread to run the actual test and storing > the > TestResult somewhere (e.g. ServletContext) with an ID. > The browser could then reload the page (automatically if required) with an > ID > for the TestResult. My only concern with this approach is any threading > issues? > > Any thoughts most appreciated. > > Cheers Dave > > > > > > ********************************************************************** > CAUTION - This message is intended for the addressee named above. It > may contain privileged or confidential information. If you are not the > intended recipient of this message you must not use, copy, distribute > or disclose it to anyone. > ********************************************************************** > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
