Re: [Scheme-reports] eq? and eqv? for records Sascha Ziemann (14 Feb 2014 16:18 UTC)
Re: [Scheme-reports] eq? and eqv? for records taylanbayirli@gmail.com (14 Feb 2014 21:40 UTC)

Re: [Scheme-reports] eq? and eqv? for records taylanbayirli@gmail.com 14 Feb 2014 21:36 UTC

Sascha Ziemann <ceving@gmail.com> writes:

> So R7RS does not guaranty that the following will always return #f:
>
> (define-record-type :identifier (identifier) identifier?)
> (eq? (identifier) (identifier))

No, it doesn't guarantee that.

(Currently it doesn't guarantee *anything* because the location-in-store
semantics for records have apparently been overseen, but that's just me
being pedantic; I think the *intent* was that the example you've given
is explicitly left unspecified as a special-case even though the rest of
the semantics is mostly obvious/intuitive.)

> In order to be sure I have to add a dontcare slot with a dontcare
> value?

You don't need to give it a value, the constructor doesn't need to take
values for all fields.  For example:

(define-record-type <token>
  (token) token?
  (dummy-slot token-dummy-slot set-token-dummy-slot!))

(eq? (token) (token))  =>  #false

Taylan

_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports