I'm starting to do the build of the new 1.3.0 release. so trunk version is 1.4.0-SNAPSHOT
I've just realised this will cause problems with hadoop trunk. won't it? I'll do a PR with hadoop-trunk that'll be critical for builds working on saturday -but it won't work until then. As for the 1.3.0 release itself, can't build it on my laptop (docker failure, sed argument problem in site doc generation). will try on an x86 system
