Recent http://download.jboss.org outage showed a critical flaw in ECFTransport design. Basically, Project Example based wizards hang waiting for http connections to timeout (it's a loooong timeout by default). Moreover, these http connections can't be aborted, which causes JBIDE-14593.

This needs to be fixed at the ECFTransport level, in order to properly support a "work offline" scenario.

So :

• a configurable timeout value should be used, when invoking the different connection methods
• canceling the download monitor should properly abort current http connection