On Saturday 11 August 2007 08:57, Audrius Meskauskas wrote: > Andrew John Hughes wrote: > > please take a look. They are available from: > > > > http://fuseyism.com/classpath/ > > > > Thanks, > > Hi, Andrew, > > Thanks a lot for testing. When looking into comparison, I have noticed > the following: > Thanks for taking the time to look them over :) > 1. It seems that there are multiple of "time out" errors like: > + Test timed out. Use -timeout [millis] option to change the timeout > value. > > Maybe we could just use that -timeout to get more precise ? > I need to double check what options the tests are run with (it's done via a script) but I think there is a timeout value set. I'll try increasing it. > 2. The socket test java.net.Socket.jdk13 failed because the port has > been already in used. This happened for me several times when the > previous tests failed to do the proper cleanup. It would be interesting > to rerun this test separately alone. > You're right, it's worth looking at this in more detail. Was this the case for 0.95? > 3. The javax.swing.JTree.TreeRobot now fails and was passing in 0.95. > This regression means that tree is not responding to the keystrokes as > required by standard. As all awt.Robot - based tests is sensible to the > keyboard and mouse manipulations when running, maybe this was a reason > of failing? > There weren't any keyboard/mouse manipulations as the tests were run remotely using xvfb (same as on builder); don't know if that could be the cause or if there were changes. Interestingly, I can't even find JTree in the ChangeLog for this year! So probably no recent changes... > Good luck > Audrius. Thanks again, -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net
Attachment:
pgpfluakmgOlh.pgp
Description: PGP signature
