Fwd: registering a charset or two

[Larry Masinter]

https://LarryMasinter.net  https://interlisp.org

---------- Forwarded message ---------
From: Martin J. Dürst <due...@it.aoyama.ac.jp>
Date: Tue, Feb 3, 2026 at 11:53 PM
Subject: Re: registering a charset or two
To: Larry Masinter <L...@acm.org>
Cc: Ron Kaplan <ron.k...@post.harvard.edu>

Hello Larry,

Very glad to hear from you. I hope this finds you well.

Registering these two charsets would certainly be a good idea. See below
for more details.

On 2026-02-04 05:21, Larry Masinter wrote:
> hi Martin,
>
> For the last few years I've been working on a project to revive and (to
> some degree) modernize the Interlisp enironment we built in the 1980s. This
> system supported the Xerox Character Code Standard (aka XCCS), with some
> modifications. XCCS was one of the main predecessors to Unicode.  The
> variations together we're calling MCCS (Medley Character Code Standard).
>
> We think it would be useful to register xccs aand mccs in the IANA registry
> of charset values.
>
> My question is now we should go about registering these legacy charsets...
> A publicaation on GitHub? An Internet Draft? There is a Xero publication
> desccribing XCCS but we'll need to prepare something for a registry entry
> for MCCS.

The actual registrations, as submitted (I assume) can be found at
https://www.iana.org/assignments/charset-reg/charset-reg.xhtml. Usually,
they refer to other documents. That's easy in many cases because most
charsets were/are documented e.g. as national or corporate standards.

For XCCS, I'd consider pointers to
https://en.wikipedia.org/wiki/Xerox_Character_Code_Standard and
https://archive.org/details/bitsavers_xeroxxnsXNNetworkArchitectureGeneralInformationMan_10024221/page/n65/mode/2up
(assuming you can verify that this information is still accurate) enough.

For MCCS, it might be enough to provide a list of additionally used
codepoints and the allocated characters, all in plain text, as part of
the registration.

The important thing is that we have enough information to convert a
sequence of octets into a sequence of characters, be that manually or
usually by implementing it as a piece of software. These days, the
"sequence of characters" is a sequence of Unicode characters. That
shouldn't be too difficult for XCCS, because I'd expect all characters
there later ending up in Unicode. For MCCS, it might be a bit more work;
I have no idea how much, you should know better.

The details for a registration are explained in RFC 2978
(https://datatracker.ietf.org/doc/html/rfc2978). These days,
registrations are quite rare, and therefore I can't tell you about
recent examples (neither successful nor unsuccessful). In any way, I
plan to be quite pragmatic and move things forward smoothly.

Please feel free to ask anything more if you still have questions.

Regards,   Martin.

> There is more context but I thought I should first check out whether we're
> pointed in the right direction.
>
> Lsrry
> --
> https://LarryMasinter.net  https://interlisp.org
>