[Aschenbath]

Summary

Fixes #793. A chained converter (fallback factories pointing at a parent converter) could not actually pick up the parent's custom hook factories for types the built-in factories already handle: the default attrs factory matches first, so the fallback is never consulted. The issue asks for a way to make a converter without the default hook factories, and this adds exactly that flag.

Change

• BaseConverter and Converter accept use_default_hook_factories=True (default, unchanged behaviour). With False, no built-in hook factories are registered for either direction, so dispatch goes to hooks registered afterwards, then to the fallback factories. A converter built this way with fallback factories pointing at a parent now delegates attrs classes to the parent's custom hooks.
• copy() now preserves the flag and the fallback factories, so a copied chained converter still delegates. This also fixes FunctionDispatch.copy_to with skip=0, where [:-0] wrongly produced an empty list.
• Docs: a note in docs/indepth.md next to the chained-converter example, and a HISTORY.md entry under NEXT.

How I checked it

• Reproduced the issue's case first: a child Converter with structure_fallback_factory=parent.get_structure_hook structured Foo as Foo(1) (the built-in attrs result) instead of the parent's Foo(101). With use_default_hook_factories=False it gives Foo(101), and unstructuring follows the parent's hook too.
• New tests in tests/test_converter.py: chained delegation (both BaseConverter and Converter), copy keeping the fallback, and user-registered hooks still winning on a no-default converter — 16 passed for the fallback/no_default selection.
• Broader suites: tests/test_converter.py, test_copy.py, test_function_dispatch.py, test_multistrategy_dispatch.py, test_baseconverter.py, test_factory_hooks.py — 96 passed.
• ruff check and ruff format pass on the touched files.

[Aschenbath]

Closing this one myself: I opened too many PRs here in a short time, which isn't fair to your review queue — sorry about that. I'm leaving #827 and #822 open. If this fix is something you'd want, I'm happy to reopen or rebase it later.