The great documentation update

We made great strides in 2.x having our documentation generated from actual code using roslyn to visit certain tests.

This issue serves as an aggregate for remaining documentation issues so we can track it in one issue and make one big last swoop to get it into its final shape.

• Our docs are flattened on elastic.co site. We closely mimic elasticsearch’s documentation tree but our generator loses much of this information. We need to be sure the documentation maintains the same tree in the generated asciidoc output.
• We do not generate documentation for the API tests yet. The idea is to visit all Usage and Crud tests and based on a consistent hash pick one of the 4 ways (sync/async and inititalizer/fluent) (was (#2239).
• similarly all our analysis tests should end up in the documentation (#1991)
• For various reasons we would like to embedd tables (BenchMarkDotNet output for instance) our AsciiDoc parser does not handle this yet (https://github.com/russcam/asciidocnet/issues/2)
• Revisit the flow of our Core docs (#2069)
• Documentation: Clarify how a Connection Pool works and that its not responsible for the actual TCP connection pooling (#2260)
• Write a dedicated section on how to see what the client is doing (fiddler, logging, tracing, debuginformation, audits) (#864 #865)
• Add Compatibility Matrix from README to docs #2243
• Document Search Template #2729
• Provide better context in .doc.cs cases see #3183
• Document Suggesters (#2928)
• Improve Parent/Child documentation (#3542)

[schmitch]

btw.

Working with documents

    Indexing
        Single Document
        Bulk
        Ingest Node
    Retrieving
    Updating
    Deleting

is completly missing

[stevejgordon]

Closing due to age/inactivity. Please let us know if this issue is still a relevant priority for your usage?