c218d875 · core f63c1e69 · built 2026-10-26 22:39 UTC? is defined for the core Result and Option only; there is no Try aspect
Gap
? is a compiler special form over exactly two types. The operand and the enclosing function's
return type must both be the core Result or both the core Option (RFC-0182), and the checker
recognises them by declaration identity, not through any aspect. A user-defined type, however
Result-like, cannot be used with ? as an operand or as a return type.
RFC-0182 accepted this closed rule on purpose, as the first two impls of a future Try aspect:
? would be defined over any type implementing the aspect, with Result and Option as its
first two implementations. That generalisation is additive, so programs written against the closed
rule keep compiling. It needs a residual/branch design (the associated types, and how the error
conversion composes with From), and it interacts with object safety (RFC-0008) and the Callable
question for closures (RFC-0161). None of that is designed or scheduled.
Impact
Only Result and Option can carry an early exit through ?. A library type that models its own
fallible or absent value must be bridged to one of the two (.ok(), .ok_or(...), or a
conversion method) before ? applies. The closed rule also fixes one semantic: None carries no
payload, so the Option row has no conversion step.
Affects
Resolution
Not scheduled. The open design question is metel-core#1637, which has no milestone. If it is taken up, RFC-0182's closed rule becomes the first two impls of the aspect rather than being replaced.