[Proof of Concept] Supportings source code extraction

[Ash258]

#3003

There is not really need to have 3 projects inside 1 repository for obvious reasons (history, size, project managment, issues / prs / board for each project, code quality of each project ...).

This is just example of pretty easy script which get all needed dependencies for project using scoop like way. This could be possible using gitmodules / trees, but not sure, how would this affect normal users.

Since current supportings are projects, which are maintained by scoop's developers, it's prety easy to extract them, create correct flow of develping, pipelining and automating release process (using appveyor on tag to test, build, publish github release, ...)

Since checkver does not return any data, updating would be done after some git status (if manifest was updated) or just assumption, that captured output would match some regex.

Feel free to suggest different way how to get, update dependencies, since this was my initial idea how to handle this.

Manifests (except yaml) are only illustrative, there is need to extract code first and create releases.

[Ash258]

Similar way is pretty common. Nice example is vscode's syntaxes. CSS extension - They just have cgmanifest, which continas github url (atom's syntaxes, textmate bundles), commit hash and run updates when needed

[niheaven]

If #3060 be merged, scoop should have version number, and this type of update (not single DLL, as a whole) should bump its number, maybe separate json is not needed😁

[Ash258]

/cc @r15ch13

Could we discuss this change? I would like to start on YAML manifests support again from Wednesday and this would make it more easier.