compiler.md

Wado Compiler

The Wado compiler (wado-compiler/) translates .wado source into a Wasm component binary. This document gives a high-level overview of the architecture; deeper topics live in their own docs:

• Optimization passes: optimizer.md
• LSP architecture: WEP 2026-04-18
• Language features: spec.md

Pipeline

Source (.wado)
  → Lex → Parse → Bind                    (per module, in loader)
  → Annotate (Analyze + Resolve + lower TIR)
  → Default-purity Check
  → Synthesis (auto-derives, template, From, serde, pre-CM effect dispatch, CM bindings)
  → Effect Check → Stores Check
  → Effect Dispatch (post-check: WithHandler / Resume desugaring)
  → Link (Package → FlatPackage)
  → Monomorphize → Erase Newtypes & Flags → Reflect Bridges
  → Lower
  → Optimize
  → WIR Build → WIR Optimize → Codegen
  → Wasm component bytes

The driver is compile_after_load in src/lib.rs.

Phase	Output	Module(s)
Lex / Parse	AST	lexer.rs, parser.rs, token.rs, syntax.rs
Bind	AST + bindings	bind.rs
Loader	All modules	loader.rs
Annotate	TIR + facts	semantics.rs, analyze.rs, elaborator/
Default-purity Check	(validation)	effect_check.rs::check_default_purity
Synthesis	TIR (extended)	synthesis/
Effect / Stores	TIR (validated)	effect_check.rs
Effect Dispatch (post)	TIR	synthesis/effect_dispatch.rs
Link	FlatPackage	link.rs
Monomorphize	FlatPackage	monomorphize/
Erase Newtypes & Flags	FlatPackage	tir.rs
Reflect Bridges (post)	FlatPackage	synthesis/reflect_bridge.rs
Lower	NirPackage	lower/
Optimize	NirPackage	optimize/
WIR Build	WirPackage	wir_build/
WIR Optimize	WirPackage	wir_optimize/
Codegen	Component bytes	codegen/

Compilation Units and IRs

Unit	Layer
Module (ast.rs)	Surface AST. Preserves source-level syntax to support wado format.
TirModule (tir.rs)	Typed IR. One per source module after annotate.
Package (package.rs)	Per-module compilation context, used from synthesis through link.
FlatPackage (flat_package.rs)	Flat list of all functions, types, and globals; used from monomorphize through the lower pipeline's planner.
NirPackage (nir_package.rs)	Normalized IR. Output of lower, input to optimize / WIR build / codegen.
WirPackage (wir.rs)	Wasm IR — closer to Wasm core instructions, used for emit-time optimization and codegen.

Codegen takes &NirPackage + &WirPackage (emit_wasm) and has no knowledge of earlier phases — the rule that keeps the back end decoupled from the front.

Frontend (per-module)

The loader runs lexer → parser → bind on every loaded module:

• The lexer extracts the optional __DATA__ section and tokenizes the rest.
• The parser builds a faithful AST. Compound assigns, comparison chains, struct shorthand, and &self parameters are kept verbatim so wado format round-trips.
• bind.rs performs local name resolution, scope/mutability checking, and use-before-define detection.

The AST is parser-immutable from this point on. The desugar-replacement surface rewrites — compound assignment (x += y → x = x + y), while / C-style for → explicit loop, for x of expr iteration (the .into_iter() / .next() dispatch and the match Some(x) => body, _ => break shape), the assert statement, the matches operator, the comparison chain a < b < c, template-string interpolations, use … namespace prefix stripping (helper::foo), and Self::method / T::method (T bound to concrete) static-call dispatch — happen inside the elaborator and are built TIR-direct: each rewrite resolves the user AST and constructs TirExpr / TirStmt nodes directly without producing synthetic AST. The implementations live in elaborator/{stmt,operators,assert,matches}.rs (resolve_while, resolve_for, resolve_iterator_for_of, resolve_compound_assign, desugar_assert, desugar_matches_expr, desugar_comparison_chain), Elaborator::strip_ns_prefix in elaborator.rs, and CalleeIdentKind / classify_call_callee in elaborator/call.rs (the prefix is resolved to its concrete type name before parameter-type lookup so argument resolution runs once with the correct expected-type hints). Synthetic call sites that need to dispatch a method on an already-resolved receiver TIR (the for-of .into_iter() / .next() calls today) reuse the AST-driven method dispatch via Elaborator::resolve_method_call_with (elaborator/method_call.rs) — that helper takes a pre-resolved receiver plus a method name and signals "no source AST" with method_id: None so no use→def edge is recorded against the synthesis site. Keeping the AST parser-shaped is what lets LSP queries land on the user's text rather than on a synthesised replacement.

Annotate (Analyze + Resolve + TIR Lowering)

semantics_of (semantics.rs) is the entry point shared by LSP and batch compilation. It runs analyze.rs for symbol-table construction and elaborator/ for type checking; bodies are then lowered into TIR.

The result, Semantics, carries the TIR modules plus an AstIndex and a use→def map (AstId → AstId, globally-unique ids). This is what makes the architecture LSP-friendly: facts are attached to AST nodes without mutating them, so cross-file navigation, hover, and rename all fall out of the same data the batch compiler uses. See the LSP section below.

The elaborator covers trait selection, generic inference, method dispatch, coercion, and effect typing. All trait calls are resolved statically — by the end of the pipeline every call targets a concrete monomorphized function. There is no runtime vtable.

One trait implemented for one receiver at several argument lists is chosen by the arguments, which are classified before they are elaborated (elaborator/synth.rs, WEP: Overload Resolution). That classification is a read-only query: it runs under Logger::quiet, and debug builds assert it recorded no fact.

Rigid and flexible type variables

The type table keeps the two apart, because a type check asks opposite things of them:

• ResolvedType::TypeParam is rigid. It stands for whatever a caller instantiates the binding item with, so inside that item it is opaque: nothing but itself is assignable to it, in either direction. let x: T = 5 in a body that declares T is a type error.
• ResolvedType::InferVar is flexible. It stands for a type the solver has yet to determine, so it accepts and records.

A rigid parameter appears only inside the item that binds it. Every use of a polymorphic signature instantiates its slots into fresh variables first (elaborator/instantiate.rs), so a callee's parameter never reaches a check as itself. Without that step the two collapse: TypeParam is interned by (name, index), so fn f<T>'s T and fn g<T>'s T are one TypeId, and a check meeting a bare T cannot tell the enclosing body's parameter from a callee's slot.

check_assignable therefore defers only what is genuinely undecided — an inference variable, a type pack awaiting expansion, an associated-type projection awaiting its impl, unknown / error — and compares a rigid parameter nominally.

Two consequences worth knowing when adding a check:

• A value is checked where its expected type is known. A callee's parameter types name its own slots until it is instantiated, so a call site instantiates first and checks the arguments once, against the substituted types. The same holds for a struct literal's fields and a parameter's default.
• A bare slot is not a constraint. struct Context<T> { fields: T } accepts whatever the literal puts in fields; that value is what fixes T. Where several values fix one slot — two fields naming it, or a sequence literal's elements — they are checked against each other instead.

A flexible variable never survives elaboration: finalize_infer_holes substitutes solved ones away and pins unsolved ones to error after reporting them, and the backend passes panic on one rather than classifying it. A rigid parameter does survive — it is what a generic body is written in — and dies at monomorphization instead.

Only the recorded facts are swept. The type table keeps every type ever considered, so a pass enumerating it selects with TypeTable::is_concrete rather than assuming the table holds only live types.

Synthesis

synthesis::synthesize (synthesis.rs) generates synthetic TIR that the user does not write:

Sub-pass	File	Output
Trait auto-derives	synthesis/traits.rs	Eq / Ord / Default for bound-driven requests (below); Inspect / InspectAlt unconditionally (total); Display for plain enums (bare case name); DisplayAlt wherever Display exists. Newtypes inherit their base's format traits at the call site (peel_transparent_newtype in synthesis/template.rs), except Inspect/InspectAlt which they override for the as Name tag
From adapters	synthesis/from_synth.rs	From impls from impl From<T> for U; declarations
Serde	synthesis/serde_synth.rs	The per-struct FieldSchema lookup, for bound-driven requests (below; body-less markers record there too). Every body is a Reflect* blanket in core:serde, not synthesis
Template strings	synthesis/template.rs	Expands template strings into Display::fmt / Inspect::inspect calls
Effect dispatch	synthesis/effect_dispatch.rs	Per-effect dispatch infrastructure for handler resolution
CM bindings	synthesis/cm_binding/	Component Model boundary adapters (lift / lower / async export)
Reflect metadata	synthesis/traits.rs	ReflectStruct / ReflectVariant / ReflectEnum / ReflectFlags impls: type_name, members, wire_name_policy, the payload / member associated tuples, and the value bridges (WEP 2026-06-13)

Synthesized impls are recorded back into the shared TraitEnv so subsequent phases query a single source of truth.

synthesis/reflect_bridge.rs is the one exception to synthesis running before monomorphize: a generic type's value bridges are named after the concrete subject and member types, so they exist only per instantiation (WEP 2026-06-13).

Bound-driven requests (WEP 2026-06-25-trait-derivation) originate during Annotate, from two funnels into the shared TypeTable::bound_driven_synth_requests set (one per TypeTable, project-wide, since each module gets a fresh Elaborator):

• A satisfied bound: Elaborator::type_implements_trait_inner (elaborator/trait_query.rs) records a (type, trait) pair whenever a T: Serialize/Deserialize/Eq/Ord/Default bound is satisfied. Eq/Ord/serde recurse through members via one walker, walk_structural_derive_members, shared with marker validation and reason-chain diagnostics; Default instead checks that every field carries a default expression (auto_derive_default_struct_type). A direct S::default() call records at static-method resolution (method_call.rs), since it reaches no bound check.
• A body-less marker: impl Eq/Ord/Default/Serialize/Deserialize for T; all validate eligibility immediately at the marker's own span (Elaborator::record_explicit_derive_request — a compile error if T isn't eligible) and then record into the same set, keyed by the target type's defining module.

The format traits (Inspect/InspectAlt/Display/DisplayAlt) are total (WEP): classify_on_bound_trait classifies them, type_implements_trait_inner short-circuits true for any type, and generation stays unconditional — so a T: Inspect/T: Display bound always holds and impl Inspect for T; markers are accepted, without gating. They record no demand request.

Two passes read the demand set as a snapshot, not a drain (consuming it would starve whichever runs second): serde_synth::synthesize_serde claims Serialize/Deserialize; traits::synthesize_traits claims Eq/Ord/Default, gating generation on set membership (SynthesisCtx::should_synthesize) instead of its former unconditional sweep.

A body-less marker is itself an Item::Impl and lands in TraitEnv's impl indexes like any other, so every gate that means "a real impl already covers this" must count only methodful impls: SynthesisCtx::has_real_impl (module-scoped, for Eq/Ord/Default generation dedup), has_methodful_impl_anywhere (module-agnostic, so a body-less format marker never suppresses the eager format body while a real impl Display for String does), and Elaborator::has_real_trait_impl_for_type (module-agnostic — a serde impl legitimately lives outside the type's module, e.g. core:serde's impl Serialize for i128).

Effect and Stores Checks

check_effects and check_stores (effect_check.rs) run after synthesis and before monomorphize. They validate that every function declares the effects and reference stores it actually requires. Synthesized CM boundary code is exempted. A separate check_default_purity runs earlier, immediately before synthesis, to gate auto-derived Default::default() bodies on pure field defaults.

Effect Dispatch (post-check)

synthesis::effect_dispatch::synthesize_post_check runs between the effect/stores checks and link. It desugars WithHandler / Resume constructs into the per-effect dispatch infrastructure (struct, mut global, dispatch wrappers). This pass is split out of the main synthesis stage so that effect-check sees the original WithHandler shape and can validate which effects are satisfied locally.

Link → Monomorphize → Erase

• link.rs merges per-module TIR into a single FlatPackage. After link, functions and types are addressed by global indices.
• monomorphize/ walks call sites, instantiates generic structs and functions with concrete type arguments, and rewrites references. Generic structs are keyed by (name, ModuleSource); generic functions are keyed by (ModuleSource, String), where the string is the bare method name for methods and a module-qualified FreeFunctionName for free functions. The ModuleSource is always the body's home module (where the template is registered in Package::functions), so same-named generics from different modules coexist. The dispatch loop panics if a queued generic has no template at its (module_source, name) key — every generic must be defined somewhere reachable (e.g. in core:prelude for built-in shapes like the tuple, whose reserved base name is []). Variadic TupleSpread nodes are expanded here. name.rs produces stable mangled names (Box$i32, identity$1, …).
• type_table.erase_newtypes_and_flags() then collapses newtypes to their base type and flag types to u32. The distinction is needed during monomorphize for trait dispatch but not afterwards.

Lower

lower.rs runs FlatPackage (TIR-shaped) → NirPackage as planner + translator: the planner ([lower::plan::plan]) runs the TIR-mutating sub-passes and produces a LowerPlan of facts; the translator ([lower::translate::translate]) is a single fold from TIR to NIR. See docs/wep-2026-05-11-nir.md.

Sub-pass	Stage	File	What it does
Pattern lowering	planner	lower/plan/pattern.rs	LetDestructure / IfLet → explicit Let + If; dense integer Match → Switch
Global extract	planner	lower/plan/globals.rs	Extracts non-constant global initializers into a per-module __initialize_module function (one per source module; disambiguated by module_source)
Boxing	planner	lower/plan/boxing.rs	&primitive / &mut primitive → Box<T> struct operations
Closure	planner	lower/plan/closure.rs	Closures → __Closure_N functor structs with __call methods; produces fn-param specialized callees; emits ClosurePlan { functor_infos }
Initialize modules	planner	lower/plan/globals.rs	Combines per-module init functions into the top-level __initialize_modules
Value copy	planner	lower/plan/value_copy/	Decides move / copy / share per consumption site (freshness, last-use, confinement, read-only-share) and inserts $value_copy$T only at copy sites; synthesizes the helpers
String collection	planner	lower/plan/string.rs	Collects literals and per-function DCE maps for the data section; emits StringPlan
TIR → NIR	translator	lower/translate.rs	Single fold over TIR producing NirPackage. Special-cases that consume LowerPlan: wide-int Match → if-else chain, Closure → ClosureToCanonical, copy_value::<T> → helper

Optimize

optimize/ runs a fixed-point loop of NIR-level passes (inlining, copy propagation, SROA, LICM, DCE, …). Local rewrites run on a worklist engine (nir_engine.rs) over the arena Body — a node is revisited only when an edit might have made it reducible — with the position-flexible rules sharing one session (optimize/peephole.rs). A per-function dirty-set gate (optimize/gate.rs) lets each pass skip functions unchanged since it last ran. See optimizer.md and WEP: NIR Optimizer Architecture.

WIR Build

wir_build/build_wir_package translates a NirPackage into a WirPackage in three stages: register types, collect function signatures, then translate each function body via FunctionTranslator. The pass is split across sibling files by concern:

File	Concern
context.rs	WirContext — accumulates types, functions, tables
types.rs	Stage 1: register TIR types as WIR type definitions
functions.rs	Stage 2: collect and register function signatures
component_plan.rs	ComponentPlan: CM-level structure (imports, exports, adapters)
translate.rs	Stage 3 driver and dispatch (translate_expr / translate_stmt / translate_block)
primitive_ops.rs	Literals, binary / unary operators, casts, array indexing
calls.rs	Function-ref resolution, builtin intrinsics, indirect calls, closure-to-canonical
pattern_match.rs	match / if let / switch lowering, variant construct / test / payload

Each helper module calls back into translate.rs for sub-expression translation; cross-module access uses pub(super) on shared fields.

Type lookup has two contracts. During registration a type may name one a later phase defines, so type_id_to_wir_type_pending yields a placeholder that the final fixup pass re-resolves. Everywhere after registration, type_id_to_wir_type treats a miss as a bug and panics: registration and lookup derive their keys through the same name::wir_*_key helpers, so they cannot drift apart without one of them being wrong.

CM canonical operations (stream / future read + write, waitable-set, error-context) carry no wir_build code: they are lowered entirely in the synthesis phase (synthesis/cm_binding/) into ordinary TIR that translates like any other function.

WIR Optimize

wir_optimize/ runs Wasm-shape-specific passes that need WIR's lower-level view: peephole, init-guard removal, struct elision, array data promotion, parameter SROA, nullable-ref folding, constant forwarding, DCE, and final cleanup. Tuple- and user-struct-return ABIs are decided by the TIR-level optimize::multi_value_return pass before WIR build, and variant returns are scalarized into tuples at NIR by optimize::sroa_variant_return; what stays here is the result-slot flattening that needs the post-nullable_ref shape.

Codegen

codegen::emit_wasm produces the final component bytes:

1. emit.rs emits core Wasm bytes from WIR.
2. component.rs wraps the core module in a Component Model envelope (imports, exports, adapters, optional WIT bundling, embedded data).
3. postprocess.rs adds branch-hint sections and other post-emission custom sections.

Output is validated with wasmparser unless --no-validate is set.

Module Loading and Names

Module Sources

name.rs::ModuleSource distinguishes where a module originated:

Variant	Origin
Core	Embedded core stdlib (core:prelude, core:cli, …)
Wasi	Embedded WASI bindings (wasi:cli, wasi:io, …)
Local	Path relative to project root (./geometry.wado)
Remote	http(s)://… URL, fetched via host.load_remote()
EntryPoint	The main file being compiled
Redirected	Module routed through a Kiln invocation index
Wasm	A .wat / .wasm asset imported via use … with { type: … }

The loader canonicalizes paths (RFC 3986, project-root-relative with / separator) so the same file imported via different paths shares one identity.

Naming Convention

name.rs centralizes mangling so other components do not depend on name shapes:

Name	Format	Example
Method	{impl}/{decl}/{Type}::{method}	./geom.wado/./geom.wado/Point::sum
Trait method	{impl}/{decl}/{Type}^{Trait}::{method}	./geom.wado/./geom.wado/Point^Display::fmt
Effect operation	{Effect}::{op}	Stdout::write_via_stream
WASI canonical	wasi:{pkg}/{iface}::{fn}	wasi:cli/stdout::write-via-stream
Mangled generic	{Base}$T1$T2…	Box$i32, Pair$i32$String

Every fq name names its subject by the module that declares it, and a type written into any name goes through TypeTable::mangle_type_arg_for_generic. A simple name alone is never an identity — two modules may declare the same one.

A method key is (impl module, declared receiver, trait, method), so {impl} and {decl} repeat whenever a type is implemented in the module declaring it — the common case. A receiver with no declaring module (a builtin, a tuple) has no {decl} segment. See WEP 2026-07-29 for why neither segment is removable alone.

The same rule binds names still in their written form. A type name in source is relative to the module that wrote it, so a reference site — not a consumer — is where an identity is derived, once: crate::resolve::Resolutions answers every site before elaboration begins, keyed by the site's own AstId. An impl block's digest (ImplHeader) carries its module, its target's ImplTargetKey and its trait's, and the whole-program checks (coherence, orphan rules, sealed traits, trait-method arity) read the digest instead of re-walking loaded_modules, because a second walk knows no module and can only compare spellings. A consumer holding a bare name with no site goes through the table's own scope lookup, so it cannot answer differently from the site. See WEP 2026-08-10.

Component Model Registries

Three registries collect declarative information from the standard library and feed both the elaborator and codegen:

• CmInterfaceRegistry (component_model.rs) — extracts WASI interfaces from lib/wasi/*.wado: version pins, async flags, canonical method names, supported types. Codegen drives import generation from this registry; only interfaces whose types are fully supported are imported.
• WorldRegistry (world_registry.rs) — collects world definitions (e.g., the Command world from wasi/cli.wado) and provides export signatures.
• BuiltinRegistry (builtin_registry.rs) — collects function signatures from lib/core/builtin.wado. Functions tagged #[canonical("ns", "name")] import a CM canonical builtin (wasi, mem, or bundled); untagged builtins compile directly to Wasm instructions.

LSP

The language server (wado-lsp/) is a thin layer on top of wado_compiler::semantics. The Engine holds open documents and answers LSP queries (diagnostics, hover, go-to-definition, references, document highlight, semantic tokens). Each query:

1. Composes parse(source) → load(parsed, …, invocations, …) → semantics_of(loaded, host, …) to obtain a Semantics snapshot (kiln invocation discovery runs against the parsed entry AST between stages).
2. Uses Semantics::cursor_at(module, line, col) → Cursor to translate a (line, col) into an AstId cursor.
3. Reads pre-computed facts off Cursor / Semantics (def_key, def_name_span, references_to_def, is_write_target, …).

Engine itself performs no I/O — every query takes an &impl CompilerHost, so the caller decides how imported modules are loaded. wado-lsp ships a FilesystemCompilerHost; embeddings (VS Code Wasm, browser playground) supply their own host. The wado-compiler crate must compile to wasm32-unknown-unknown to support those bundled deployments; CI enforces this. See WEP 2026-04-18: LSP Architecture.

Kiln (Schema-Driven Code Generation)

Kiln is the compiler's mechanism for turning external schemas (.proto, .graphql, .g4, .wit, custom IDLs) into .wado source. A generator is itself an ordinary Wado package that targets the core:kiln/generator world; the compiler builds it to a Wasm component, and the host (wado-cli) executes it via wasmtime to produce .wado source files. Invocations are content-addressed by their schema bytes, options, and the generator's source hash. See WEP 2026-04-12: Kiln.

The compiler-side pieces live in src/kiln/:

Module	Concern
invocation.rs	Canonical Invocation representation (declaration site + options + schema paths)
inline.rs	Collects inline use … with { generator: … } invocations (InvocationIndex)
plan.rs	DAG + topological sort of invocations; rejects cycles
cache.rs	Cache-key composition over schemas, options, and the generator's identity hash
header.rs	Generated-file #![generated] header emission and parsing
metadata.rs	Persisted per-invocation cache state (<primary>.kiln.json)
options.rs	Extracts an OptionsDescriptor from a generator's pub struct Options
options_check.rs	Validates user-supplied options against the descriptor
import_check.rs	Refuses wasi:* imports inside generator packages; injects the Request<T> adapter

The pipeline driver (run_generator, file persistence, cache-state handling) lives in wado-cli; wado-compiler exports only the pure-data pieces above so it stays wasm32-unknown-unknown-clean.

compile_after_load integrates Kiln at three spots:

• Phases 1a–1b (pre-analysis): when the entry module's target world is core:kiln/generator, import_check rejects forbidden imports and rewrites fn generate(req: Request<Options>) into the wire shape — a primary parameter, an inputs parameter, and a typed options: Options parameter — rebuilding req as a local so the author's body is unchanged.
• Phase 6c (post-annotate): if the target world is core:kiln/generator, extract_options_descriptor walks the resolved pub struct Options and produces the descriptor the CLI provider caches on disk, so a later run type-checks a use site's options without recompiling the generator.
• Module loading: when an import target matches an entry in the caller-supplied InvocationIndex, the loader resolves it to ModuleSource::Redirected { uri } and asks the host to load the generated bytes. This is how use Foo from "schema.proto" with { generator: ... } gets wired up after generation.

Standard Library

The compiler bundles the standard library inside its binary (stdlib.rs embeds lib/):

• lib/core/ — prelude, cli, collections, serde, json, simd, zlib, base64, url, router, kiln, internal, builtin, allocator, plus co-located _test modules.
• lib/wasi/ — Wado bindings for WASI P3 interfaces. Generated from WIT by wado-from-idl; regenerate with mise run update-stdlib-wasi.

Bundled Math (wado-bundled-libm)

The wado-bundled-libm/ crate compiles a deterministic libm to wasm32-unknown-unknown, checked in as lib/core/libm.wat (rebuild with mise run update-bundled). core:prelude name-imports its exports through the ordinary core-wasm asset import (use { libm_sin as f64_sin, … } from "../libm.wat" with { type: "wat" }) and attaches them to f32 / f64. The compiler links the asset as a separate core module inside the produced component, pruned to the exports the program actually calls.

Allocators

Three allocators live in lib/core/allocator.wado, each tagged #[allocator("name")]. The compiler picks one by setting that function's export_name to "realloc":

Mode	Default for	Behaviour
bump	CLI	Bump pointer with free-rewind; never reclaims general blocks.
freelist	HTTP service worlds	First-fit free list with block splitting; falls back to bump.
debug	Test world / E2E tests	Never reuses freed memory; poisons with 0xFF. Catches use-after-free.

--allocator <name> overrides the defaults.

In Progress

• Variant pattern matching: struct payloads not yet supported (single-payload and tuple-payload work).
• Function types: parser supports fn(T) -> U and closure codegen works, but full first-class function types are incomplete.
• Stream/Future: resource declarations exist in core:prelude/types.wado, but method resolution (.new(), .read(), .write(), .close(), .drop()) is still hardcoded in elaborator/method_call.rs rather than driven by the resource declarations.

Known Limitations

• Implicit struct literals do not work with generic structs: let b: Box<i32> = { value }; fails. Use let b: Box<i32> = Box { value };.
• GC arrays cannot be passed directly to stream<u8> — they must be copied to linear memory first (component-model#525).
• A very long || / && chain aborts the compiler with a stack overflow. It parses as a left-nested binary expression, so every recursive expression walk descends once per operand; 2400 operands need ~256 MiB of stack (measured with ulimit -s — the 64 MiB wado-cli asks tokio for is not enough). Repro: a fn returning c == 0 || c == 1 || … with 2400 terms. The pattern path takes the same width, so Gale emits c matches { … } for its lexer dispatch.

Not Yet Implemented

• ? operator (error propagation)
• Effect handlers
• Reactive signals (source values, derived values, effect blocks)