Not sure what might be causing the SAX problem. Can you file an issue with more detailed information? The NoClassDefFound is caused by not copying gwt-servlet.jar into WEB-INF/lib. If you run the ant build file, it will do this automatically, but we have a fix in place for the next milestone that makes webAppCreator go ahead and do this for you for the case where you're just loading the project up in Eclipse without ever running the build file.
This particular error state is now detailed here in the "common pitfalls" section: http://code.google.com/docreader/#p=google-web-toolkit-doc-1-6&s=google-web-toolkit-doc-1-6&t=DevGuideImplementingServices On Fri, Feb 13, 2009 at 2:25 PM, Jeff J. <jsjenkins...@gmail.com> wrote: > > I'm getting class not found exceptions when trying to launch my > project in hosted mode using this build, saying Provider > org.apache.xerces.jaxp.SAXParserFactoryImpl not found. I tried to > create a fresh new project using webAppCreator and I get > NoClassDefFoundError looking for com/google/gwt/user/client/rpc/ > RemoteService. I'm using jdk1.6.0_11 32bit JVM on windowsXP 64bit. Any > suggestions on what I could be doing wrong? For now I am needing to > use r4603 from the trunk with the modified jetty jar in the gwt tools > to use hosted mode. > > Thanks, > Jeff > > On Feb 6, 3:22 pm, Scott Blum <sco...@google.com> wrote: > > On Fri, Feb 6, 2009 at 12:30 PM, Matthew Mastracci < > matt...@mastracci.com>wrote: > > > > > 2. The compiler task crashes Java 1.6 on Mac now. The previous > versions > > > would run fine with the 64-bit JVM (even though SWT isn't compatible). > > > Would it be possible to lazy load SWT so that we can take advantage of > the > > > faster 1.6 JVM on OSX? > > > > > [java] Invalid memory access of location 00000000 rip=01160767 > > > > Good point, I can take a look at this. The compiler should be able to > run > > without loading SWT. > > > > 5. One cool thing that was missing from the release notes is new the > > > > > > > > > ability to specify multiple modules for the compile task. Based on > some > > > wallclock testing, specifying multiple modules shaves off ~30 seconds > off > > > compile time of any modules on top of the first: > > > > > Individual times: > > > > > [java] Compilation succeeded -- 51.814s > > > [java] Compilation succeeded -- 170.986s > > > [java] Compilation succeeded -- 91.488s > > > > > All together: > > > > > [java] Compilation succeeded -- 55.121s > > > [java] Compilation succeeded -- 141.052s > > > [java] Compilation succeeded -- 55.492s > > > > > Note that each module's log messages are indented one level further, > > > indicating the extra speed boost. ;) > > > > LOL! I had no idea this actually made any difference in speed, so I > guess > > it's a bonus easter egg feature. :P > > > > Scott > > > --~--~---------~--~----~------------~-------~--~----~ http://groups.google.com/group/Google-Web-Toolkit-Contributors -~----------~----~----~----~------~----~------~--~---
