Motivation and Context
As webpack 1 is now officially deprecated, we are upgrading to webpack 2. The initial idea was to reduce asset size by employing tree shaking, but at the current state there is only a theoretical benefit as CommonJS imports are not statically analyzable and therefore unaffected by it. Right now this PR is only doing the required changes to be able to use webpack 2 (slight changes of the config syntax, removal of some now obsolete config parts, some minor changes of TS imports/exports due to stricter semantics) and updates of webpack/babel version numbers, no further optimization has been performed yet.
This is currently in progress, as it would require changes to plugin webpack configs.
How Has This Been Tested?
Web Interface and plugin assets have been bundled and quickly tested.
I tested executing the production build (using the graylog project and executing a