I've just published draft #3 of SRFI 136.  It was submitted by Marc

Nieper-Wißkirchen, the author of the SRFI.

  http://srfi.schemers.org/srfi-136/

Here's the diff:

  https://github.com/scheme-requests-for-implementation/srfi-136/pull/2

Here are Marc's notes on this draft:

  This draft incorporates two important changes:

  The concept of a runtime record-type descriptor is being introduced.

  In previous iterations of this proposal, the predicate procedure of

  a record type served a second purpose, namely that of a runtime

  representaton of the record type descriptor to be used in the

  procedural interface.  This led to some misunderstandings and a

  couple of questions raised in

  http://srfi-email.schemers.org/srfi-136/msg/4000979.  These

  questions have been addressed in this modification of the proposal,

  which introduces the notion of a runtime record-type descriptor,

  which can be extracted from the (syntactic) record-type descriptor

  and which frees the record-type predicate from having to serve two

  purposes.

  The implementation of SRFI 136 is now being based on top of SRFI 137

  (whose sample implementation has been included in the repository).

  This is mainly in order to demonstrate how to use SRFI 137 and how

  SRFI 136 could be implemented on the primitives provided by SRFI

  137.

Regards,

SRFI Editor