ABNF guidance

All,

(This note is being sent on behalf of the W3C Internationalization Core 
Working Group, which I happen to chair. Note that this also means that 
it is as a technical contributor and not as editor of 4646bis)

During our most recent teleconference [1], the I18N Core WG discussed 
the changes proposed for the ABNF by LTRU (to remove the extlang 
production).

I was tasked with reporting some concerns about how to address this with 
Spec writers at W3C and elsewhere. The concern is that the I18N Core WG 
has been recommending that references to RFC 3066 be changed to 
reference BCP 47 (rather than a specific RFC number). In one particular 
case (XML Schema 1.1) we requested that their language production be 
derived from the ABNF in BCP 47, that is, from the ABNF in RFC 4646.

The problem is that productions included into XML DTDs are then relied 
upon by implementations and difficult to change. There is naturally some 
resistance to changing the production in XML Schema and some concern 
that the grammar isn't fully stabilized. There is thus concern about the 
future stability of the ABNF. The I18N Core WG would therefore like to know:

- if LTRU can make a stability statement about the ABNF in 4646bis (e.g. 
what may change in the future and what is unlikely to change)
- if LTRU could include text about compatibility between 4646 ABNF and 
4646bis ABNF in 4646bis
- if LTRU would recommend a particular reference to an RFC or continue 
to recommend the BCP reference for this purpose

Best Regards,

Addison

[1] http://www.w3.org/2008/01/16-core-minutes#item06
-- 
Addison Phillips
Globalization Architect -- Yahoo! Inc.
Chair -- W3C Internationalization Core WG

Internationalization is an architecture.
It is not a feature.

Received on Thursday, 17 January 2008 21:56:15 UTC