Two suggestions about record identity Alexey Radul 13 Aug 2008 03:25 UTC
The current wording seems to prevent an implementation from making
equal? a generic function which could have user-specified behavior on
certain records (while perhaps defaulting to mimicking eqv?), whereas
I would assume such an implementation would be desirable. Please
clarify as to whether you intend to permit something like that.
Perhaps this can be accomplished by the following:
"Implementations are permitted to provide mechanisms (such as a
generic operation facility) for users to customize the behavior of
equal? on records. In default of the exercise of such mechanisms, two
ERR5RS records are equal? if and only if they are eqv?."
Also, please specify (or at least explicitly leave unspecified)
whether two ERR5RS records without fields are eqv?.