This PR makes loading cached data parallel to loading source. This turned out to be faster than keeping sources and cached data in a single blob and its also faster than the current waterfall approach. This work is for https://github.com/microsoft/vscode/issues/62643
This PR makes loading cached data parallel to loading source. This turned out to be faster than keeping sources and cached data in a single blob and its also faster than the current waterfall approach. This work is for https://github.com/microsoft/vscode/issues/62643