docs(protocols): second whole-repo pass over link and transport (#1103)

[JarryShaw]

Please follow the guide below

• You will be asked some questions, please read them carefully and answer honestly

• Put an x into all the boxes [ ] relevant to your pull request (like that [x])

• Use Preview tab to see how your pull request will actually look like

• Searched for similar pull requests

• Followed the coding style (make pylint, make mypy, make isort)

• make test passes, and a test case covers the change. Narrow runs only, listed below.

• Added a changelog entry under docs/source/changelog/ and regenerated CHANGELOG.md, if the change is user-visible: N/A — centralised in docs(changelog): shared 1.5.0 changelog — long-lived, merges last (#610, #616, #617, #618, #620) #657

What is the purpose of your pull request?

Tick the commit type your subject line carries.

• fix — corrects a defect
• feat — adds a feature
• perf — changes performance, not behaviour
• refactor — changes neither behaviour nor performance
• test — tests only
• docs — documentation only
• ci — workflows or build tooling
• release — bumps the version or rolls up a distribution
• chore — anything else

---

Summary

#1103 second whole-repo pass, linktrans slice.

Backfills the 1.1.0-1.4.1 version directives for the link and transport
public API, each checked against the release tags it names.

1.1.0, TCP data model and header schema (v1.0.3 against v1.1.0):

• Timestamps: versionchanged, renamed from Timestamp; the
  TCP_Timestamp package export became TCP_Timestamps.
• PartialOrderServiceProfile: versionchanged, renamed from
  PartialOrderConnectionProfile, with its TCP_ export renamed too.
• schema MPTCP: versionchanged, an EnumSchema keyed on MPTCPOption
  rather than an Option subclass; Option registers _MPTCP instead.

1.3.0, next-layer registration (v1.2.2 against v1.3.0):

• Link.register and Transport.register: versionchanged, one protocol
  argument (descriptor or class) replaces module and class_.

Rendered pages for the slice show no raw markup and no build warnings.
ast.dump is identical with docstrings blanked on all four files;
tests/project and the 34 test files that touch these paths pass.

Refs #1103

[JarryShaw]

Verdict: GOOD TO GO at 711e7c836. A Sonnet cross-review (author: Opus) found no blocking findings. Scope stays inside the slice, the AST guard is identical against main, and every directive was checked against the release tags. The merged build of all nine second-pass heads has 18 warnings, the same as main. tests/project: 950 passed, 1 skipped, 3 warnings, 3670 subtests passed.