Track WebAssembly/component-model#686 (web-embedding draft): post-A10 alignment watchlist

[lannbot]

WebAssembly/component-model#686 is eqrion's draft web embedding for components (binds components directly to WebIDL, replacing the Explainer's JS-API sketch; Gecko prototype behind it). It is the presumptive shape of eventual native support, so divergences between it and our embedder API are migration surface.

This issue is tracking only. The PR explicitly states it needs more iterations and is not ready for detailed review — no upstream feedback goes out from here. Revisit when the author calls for review or the draft visibly stabilizes.

Amendment A10 (#114) already took the two cheap alignments: ComponentException naming and { kind, value } variant dictionaries.

Watchlist

#	Surface	deltic (embedder-api.md)	#686 draft	Disposition
1	option-typed record fields, none case	field absent on lift	option<T> → T? as required member → null present	cosmetic-ish; would touch lift shape + bindgen types. Wait.
2	exception inheritance	ComponentException extends Error + structured payload	ComponentException derived from DOMException (TODO in draft)	DOMException absent in bare sm/jsc shell lanes; a payload-carrying subclass would stay brand-compatible if we ever move. Wait.
3	instance import/export object shape	two-level, interface-name-keyed records	TODO upstream	the only item with structural-change potential for us. Watch closely.
4	s64/u64	bigint	long long (JS Number; BigInt throws at the JS binding)	hold ours; expect the draft to move.
5	list<u8> / numeric lists	Uint8Array (always a copy)	sequence<octet>; no BufferSource coercion path	hold ours; expect the draft to move.
6	own disposal	[Symbol.dispose]/drop() + FinalizationRegistry backstop	FinalizationRegistry only	hold ours.
7	streams/futures	Stream<T>/Future<T> handles (cancel/drop expressible)	ReadableStream?/Promise — TODO in draft	nothing to align to yet.
8	variant dictionary	{ kind, value } (adopted, A10)	{ kind, value }	aligned; residual risk is upstream churn away from it (re-migration accepted — contract A10 note).

Re-check triggers

• New iterations pushed to the PR (author: "needs a couple more iterations") or the PR merging.
• The Mozilla prototype landing/shipping in Nightly.
• WebIDL-spec coordination artifacts appearing (anonymous/non-global namespaces, per-instance interfaces).

Refs: contracts/embedder-api.md (A10 status entry + terminology note), PR #114.

[lannbot]

Empirical input for the watchlist, now merged (#117): the zero-glue platform-class pattern is documented (contracts/embedder-api.md §Resources, non-normative subsection) and pinned executable (runtime/tests/embedder/platform_class_test.ts). Two pins bear directly on rows here:

• Row 1 (null vs undefined/absent): pinned — a native returning null where WIT says option<T> is NOT none; the call fails with the conversion layer's TypeError naming the import. The draft's nullable mapping (option<T> → T?) would accept it. This is the concrete interop cost if upstream keeps nullable and we keep undefined-for-none; the one-line bridge (?? undefined) is documented.
• Row 3-adjacent (instance imports): the pattern exercises our two-level interface-keyed imports shape against tagged [constructor]/[method] imports — whatever shape upstream's instance-import TODO lands on, this test is where the delta will surface first.

Also pinned for future wat authors: wasmparser's all_valtypes_named_in_func requires records referenced by imported functions to be named type exports of the importing instance.