Re: Registering a charset alias

[Anne van Kesteren <annevk@opera.com>]

On Thu, 13 Aug 2009 20:54:04 +0200, Erik van der Poel <erikv@google.com> wrote:
> I agree that it would be great to record the "de facto" standard
> charset names somewhere. Preferably at IANA, where the "de jure"
> charset names are registered.
>
> Normally, IANA does not register charset names that start with "x-",
> but since they are accepted by major implementations, it would be nice
> if they were recorded somewhere. The "x-" names should probably only
> be recommended for input, however. For output, we should strongly
> recommend the names without "x-", preferably a single name for each
> character encoding.

Yeah, I believe HTML5 currently encourages authors to use the preferred IANA name of the encoding.


> It would also be great if we could record the actual superset
> relationships that major implementations use. For example, when a
> document is labelled with the charset gb2312, major implementations
> use the "superset" gbk instead.

Just like US-ASCII or ISO-8859-1 is treated as windows-1252? HTML5 includes these mappings too, currently in section 2.7, but we were not sure whether IANA would formally register them. If that can be done I think that would be awesome.

If it can only be done for a few of the encodings, but not all, because e.g. we think that keeping ISO-8859-1 and windows-1252 distinct for some usage is desirable, that would be OK too. The more is dealt with at the registry level the better in my opinion.


-- 
Anne van Kesteren
http://annevankesteren.nl/