Record inheritance bug?

[Derick Eddington]

Just noticed this, no priority to me. If my interpretation of the report is correct, I think this should make B be a subtype of A:

Ikarus Scheme version 0.0.3+ (revision 1537, build 2008-07-13)
Copyright (c) 2006-2008 Abdulaziz Ghuloum

> (define-record-type A)
> (define-record-type B
    (parent-rtd
      (record-rtd (make-A))
      (make-record-constructor-descriptor (record-rtd (make-A)) #f #f)))
> (A? (make-B))
#f ;; Should be #t right?
>

Doing the analogous thing (at least, the most analogous thing I could figure out) with the procedural layer works:

>
  (define C-rtd
    (make-record-type-descriptor 'C #f #f #f #f '#()))
  (define C-cd
    (make-record-constructor-descriptor C-rtd #f #f))
  (define make-C
    (record-constructor C-cd))
  (define C?
    (record-predicate C-rtd))
>
  (define D-rtd
    (make-record-type-descriptor 'D (record-rtd (make-C)) #f #f #f '#()))
  (define D-cd
    (make-record-constructor-descriptor
     D-rtd
     (make-record-constructor-descriptor (record-rtd (make-C)) #f #f)
     #f))
  (define make-D
    (record-constructor D-cd))
>
> (C? (make-D))
#t
>

[Abdulaziz Ghuloum (aghuloum)]

Fixed in 1538. define-record-type simply ignored the parent-rtd clause.