Glyn Normington wrote: > > I can't understand why this build failure crept in, but I'm sure it was due > to the changes I made yesterday. I backed out the relevant change (adding a > SOAPException to the no-args constructor of the Axis SOAPEnvelope class) > this morning UK time, but Gump appears to have run earlier than that. > > What I can't understand is that I did a "cvs update -d2 of my whole Axis > tree (xml-axis and its contents) and a successful "ant clean all" before > committing my changes yesterday, so I don't understand why > SignedSOAPEnvelope didn't fail to compile for me. Any thoughts on this > mystery? If I can't trust this technique to prevent introducing build > breaks, can anyone suggest a better technique?
It is quite possible that you compile without including XML security in your classpath, as it is both optional and causes the tests to take an inordinant amount of time. Anybody know why we rerun *ALL* of the unit tests with security turned on, if it is available? To me this seems like gross overkill... - Sam Ruby
