Swap branches and please go with v0.7 from now

[Luke Han]

Hi all,

    After several months development and test, we would like to introduce Kylin's new design from Inverted-Index branch (v0.7.x). The new one contains many metadata changes, huge bug fixes, job flow refactoring, design enhancement, GUI refactoring, new deployment, the coming streaming support also will fork from this branch, the first Apache Kylin release will also start from v0.7.x. 

    Since it's more stable and will be our primary source, we will promote this branch to "master" and rename current master to "v0.6" branch in one or two days, the "staging" branch will also be v0.7.x.

    Please go with this new version (v0.7.x) from now, you can deploy from source code or with binary package. The binary package deployment no longer requires maven/npm/bower and others on your destination server, it's only depends on your Hadoop environment. We believe this deployment approach will reduce a lot of problems many of you have faced previous. 

    For binary deployment, please refer to download page here: http://kylin.incubator.apache.org/download/. Please check more detail about the installation guide especially you want to deploy on same machine which running v0.6 version. 

    IMPORTANT NOTES: 

    1) The old metadata from v0.6.x is not easy to migrated to new one, our suggestion is to create cube again through new GUI when you switch to new version, and rebuild cubes. Please let's know if there's any problem during cube creation or build, we will figure out how to deal with them.

    2) If your local code repo is cloned from previous master branch, please do not run "git pull". We recommend to clone again from new master code. Yang will brief more detail later after we successfully switch branches.

    3) Backup all your metadata/configuration before upgrade.

     Please let's know if there's any issue.

Thanks.