Re: [Ltru] Ticket #45: updated editor's copy available

[Mark Davis <mark at macchiato.com>]

Doug had a small fix, for the use of macrolanguage subtag. So

This form
may be useful in environments where the presence of the macrolanguage
subtag is considered beneficial in matching or selection (see <xref
target="choiceUsingExtlang"></xref>).
=>
This form
may be useful in environments where the presence of the Prefix
subtag is considered beneficial in matching or selection (see <xref
target="choiceUsingExtlang"></xref>).

Also fixed on http://www.macchiato.com/unicode/ltru-canonicalization

Mark

On Wed, May 6, 2009 at 07:42, Mark Davis <mark at macchiato.com> wrote:

The other formulation we could use goes back a bit more towards the v21 text with a single canonical form, yet still have a clearly defined extlang form for people who need it.  If you can use the web, it is a new third column on http://www.macchiato.com/unicode/ltru-canonicalization. The differences from the last formulation (B) I sent out are:

[Remove from B: "There are two canonical forms... ...considered beneficial in matching or selection."]

[Replace B: "4. In the extlang canonical form (but not...
...Step 3." by the following]

The canonical form has no extlang subtag. There is an alternate 'extlang form' that modifies the canonical form so that primary language subtags that are also extlang subtags are prepended with the extlang's Prefix. This form may be useful in environments where the presence of the macrolanguage subtag is considered beneficial in matching or selection (see <xref target="choiceUsingExtlang"></xref>).

• For example, "hak-CN" (Hakka, China) has primary language subtag 'hak', which in turn has an 'extlang' record with a Prefix 'zh' (Chinese). The extlang form would be "zh-hak-CN" (Chinese, Hakka, China).

I could live with either formulation.

Mark

On Tue, May 5, 2009 at 23:57, Randy Presuhn <randy_presuhn at mindspring.com> wrote:

Hi -

As a technical contributor...

> From: "Phillips, Addison" <addison at amazon.com>
> To: "LTRU Working Group" <ltru at ietf.org>
> Sent: Tuesday, May 05, 2009 11:07 AM
> Subject: [Ltru] Ticket #45: updated editor's copy available
...

>    There are two canonical forms for language tags: the 'default'
>    canonical form contains no extended language subtags, while the
>    'extlang' canonical form contains extended language subtags where
>    required.  Normally, the 'default' canonicalization is preferred.
>    However, the 'extlang' canonical form can be useful in environments
>    where the presence of the enclosing primary language subtag is
>    considered beneficial to matching or selection (see Section 4.1.2)

...

The word "required" looks a lot like the RFC 2219 keyword "REQUIRED", which
is essentially a "MUST".  Reading the proposed text in that light, the
"default" form makes no sense.

I really really really really really dislike, loathe, and abhor the very idea
of having two "canonical" forms - it seems oxymoronic.  Yes, there are
two use cases, but that does not mean that both have to be called
"canonical", or that the one true canonical form (whichever that is)
needs to serve both purposes.

Randy

______________________________ _________________
Ltru mailing list
Ltru at ietf.org
https://www.ietf.org/mailman/listinfo/ltru