Re: Fixing the IANA charset registry

[Bruce Lilly <blilly@erols.com>]

On Fri December 3 2004 13:20, Paul Hoffman / IMC wrote:

> If anyone finds bugs in the IANA charset registry, feel free to send 
> them to me directly. I'll organize them, get them discussed on the 
> list, and then bring a set of well-formed requests for changes to 
> IANA.

OK; I'm copying the mailing list so that you don't get hundreds of
responses pointing to the same bug(s), and I've set Reply-To to
point to the list in case there's any discussion warranted.

1. Alias "ASCII" is a violation of RFC 2046 which specifically reserved
   (meaning it was not ever supposed to be used) that name.

2. Several entries have lines containing
    Alias: none
    which is confusing and silly. If there are no aliases, simply don't
    provide an Alias line [but see #5 below].

3. Several charsets have inappropriate MIBenum values. I discussed
    several then-recent ones in my message to this list at the end of
    January of this year. And of course there's ISO-10646-J-1 which
    has no MIBenum value at all.

4. Lack of indication of suitability for MIME text use (also mentioned
    in January).

5. Many charsets are missing cs* aliases as required by RFC 2978
    section 2.3. (Another issue mentioned in January)

6. [this is really a side-issue] Some of the references have gross errors
    or are otherwise out of date (e.g. RFC 1345 which has a number of
    errors).