Re: Registering a charset alias

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

Hi Ira, thanks for your response!

On Fri, 14 Aug 2009 01:14:51 +0200, Ira McDonald <blueroofmusic@gmail.com> wrote:
> IANA won't register a free-standing charset alias (there's no procedure),
> but you can revise an existing charset registration.

Ok.


> So:
> (1) You could NOT update the 'Windows-31J' and 'Big5' registrations
>       to add 'x-sjis' and 'x-x-big5 aliases, because of the 'x-'  
> prefixes.

What exactly do we need to do with these then? Handle them outside the IANA registry as they are handled now? Is that ok? Given that there is content out there that uses the "x-" prefix and relies on them working would it not make sense to register them anyway. We could indicate for instance they are not to be used when producing new content.


> (2) You could update 'Windows-31J' for a 'Windows-932' alias.

Cool!


> (3) You could NOT update the 'Windows-1252' registration to add
>       'US-ASCII' (the preferred MIME name of ANSI_X3.4-1968) or
>       'ISO-8859-1' (the preferred MIME name of ISO_8859-1:1987),
>       because they would be ambiguous names.
>
> Does HTML5 *really* make such an unwise suggestion about
> treating US-ASCII and ISO-8859-1 as Windows-1252?

The reason this is done is because there's a vast body of content out there that relies on these mappings and user agents have supported these mappings for ages. (Especially ISO-8859-1 to Windows-1252.) There is no way that we can see that we can undo this.


It seems not all feedback on character encodings has been incorporated into HTML5 just yet. Here's more information on legacy mappings and legacy encodings:

  http://www.mail-archive.com/whatwg@lists.whatwg.org/msg16301.html


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