HTTP semcomv depends on other experimental docs

[lmolkova]

HTTP semconv explicitly or implicitly depends on the following documents with experimental status which should(? partially?) be marked as stable before HTTP semconv can be marked as stable:

• span_general.md. In particular, the following attributes:
  • net.peer.name
  • net.peer.port
  • net.host.name
  • net.host.port
  • net.sock.family
  • net.sock.peer.addr
  • net.sock.peer.name
  • net.sock.peer.port
  • net.sock.host.addr
  • net.sock.host.port
  • net.transport Remove mention of net.transport from http semantic conventions #3244
  • Confirming it's ok to mark HTTP semconv stable without marking net.* attributes stable #3260
• attribute_naming.md. If this doc changes significantly, we'd need to rename HTTP attributes
  • [UPDATE] per @arminru comment below, changes in this doc should not require changes in stable conventions
  • [UPDATE2] Mark Attribute naming conventions as stable. #3220
• attribute-requirement-level.md. If we change the definition of requirement levels, this could cause breaking changes for end-users
  • [UPDATE] Attribute Requirement Levels needs to be marked as stable #3235
• telemetry_stability.md - the definition of telemetry stability is experimental.
  • [UPDATE] Mark telemetry schema readme stable. #3221
• schemas/README.md. Even though versioning-and-stability.md is stable, it relies on schema transformation which is experimental.
  • [UPDATED] Provide clear guidance on what semconv actually enforces #3225

[arminru]

Thank you for collecting these dependencies, @lmolkova!

[arminru]

attribute_naming.md. If this doc changes significantly, we'd need to rename HTTP attributes

I don't think this necessarily blocks stabilizing the HTTP semantic conventions.
If changes were made there, we could call out that existing, stable conventions will remain as-is.

[trask]

@lmolkova thx for putting this list together! do you think the remaining stability dependencies are now tracked by

• Confirming it's ok to mark HTTP semconv stable without marking net.* attributes stable #3260
• Attribute Requirement Levels needs to be marked as stable #3235
• Provide clear guidance on what semconv actually enforces #3225

and if so are you ok with closing this one?

[lmolkova]

closing in favor of ⬆️