Algolia crawler requires the baseURLs be bare domains, otherwise a conflict arises between the loaded page and the defined canonical URL in each page. This causes the Algolia crawler to reject the page and not find anything to index in the docs.
Also corrects Intersphinx errors at build time resulting from not having the correct paths to linux and kubernetes docs.
This commit also generally cleans up the Installation docs and uses /includes/<platform>/steps-*.rst as a pattern for storing platform-specific tutorials within a single root URL.
Finally, due to API rate limits, this commit adds some Makefile magic to allow skipping of sync'ing dependencies. Pass `make BUILD_DEPENDENCIES=FALSE <platform>` to skip building any deps and use what is already in the branch.
Klaus suggestions
Adding Eco's suggestions
Co-authored-by: Eco <41090896+eco-minio@users.noreply.github.com>
CR updates
Additional review comments
final work
