Re: [Ltru] [apps-discuss] Fwd: Defining a CBOR tag for RFC 5646Language Tags

[Dave Cridland <[e-mail]>]

On 14 May 2014 04:33, Peter Occil <[e-mail]> wrote:

> I'm not aware of any use case where having multiple language tags for the
> same plain-text string is useful.  For instance, RDF supports only one
> language tag for each string.  And HTML5 doesn't support multiple languages
> in the Content-Language header field or META tag; instead, for multilingual
> documents, it relies on markup to set the language used for each section.
> But plain-text strings don't admit of HTML-like markup without more.
>
> Moreover, having multiple language tags for plain text leads to the
> additional problem of determining which parts of the text each language tag
> applies to, which is not so easy in the case of your three-language example.
>

Many years ago, Mark Crispin and Chris Newman had a proposal for embedding
language tags in invalid UTF-8; I seem to recall they publicly renounced
their proposal rather dramatically in favour of a Unicode Consortium
proposal for embedding the language tags somewhere in Plane 14 - published
as RFC 2482.

The fact it was all initiated in order to support the pressing needs of
ACAP might give you some hints as to why it never really took off, but as a
counter-proposal to language tags in metadata, it might be worth
re-examining.

Dave.