This proposal suggests to use the IANA Character Sets Registry.
I had some concerns about the state of the registry, but I have been
able to register the UTF-7-IMAP charset recently. It may have helped
that it was specified in RFC 3501 in 2003 (but the corresponding IANA
action had been omitted), and it is still in wide use today, without a
deployed alternative. However, the basic process appears to be working,
including MIB assignment.
I think it's a great thing knowing that extending the registry is still possible.
The primary reason it's not extended often seem to be that there are few encodings
used for interchange that are not yet registered in it - and new encodings are rare
as utf-8 negates the need to come up with new encodings.
The reason we are going with this list and not something else is that:
To the best of my knowledge, it is by far the more complete one -
iso encodings are registered with IANA while the opposite isn't necessarily true.
Notably, it was important that the list contain ebcdic encodings and encodings that are used
in most environments that have a C++ compiler and/or can run C++ programs
I can't comment on the feasibility from a ISO process perspective of
using the IANA registry. ISO has its own character set registries,
somewhere in the context of ISO 2022 and ISO 4873. It may be a
nonnegotiable requirement that an ISO standard uses an existing ISO
From a wording perspective, IANA is not relevant - and not mentioned.
Instead, we refer to the RFC that describes the registry which describes
how the IANA registry works. There is precedent in the standard to mention RFCs.
I hope that answers your questions, let me know if you still have concerns!
Red Hat GmbH, https://de.redhat.com/ , Registered seat: Grasbrunn,
Commercial register: Amtsgericht Muenchen, HRB 153243,
Managing Directors: Charles Cachera, Brian Klemm, Laurie Krebs, Michael O'Neill