Re: [Ltru] Uniqueness of variant subtags

Actually no. What we want to do is register a small number of variant tags that can be productively combined. For example, one subtag 'ungegn' applies to some 47 languages that they have romanizations for. Registering each one of these would be, of course, a nightmare, when a single variant tag would suffice.

And I see no reason why 'ungegn' would be a problematic registration, and somehow require an extension. It is just a way to express an orthography, and we have many of those already. I'm just saying that we could provide some guidance in the wording so that we don't get into interminable, unproductive, case-by-case, brou-hahs in the languages group.

Mark

On Thu, Oct 2, 2008 at 5:09 PM, Phillips, Addison <addison at amazon.com> wrote:

You're planning to register "dozens" of romanizations?

I liked the idea of an extension for transcriptions much more than using variants. If it's just a few transcriptions, that's one thing. But we could fill the registry with various romanization schemes (and then there are other to-non-Latin script transliterations to consider). Ick.

Addison

Addison Phillips

Globalization Architect -- Lab126

Internationalization is not a feature.

It is an architecture.

From: mark.edward.davis at gmail.com [mailto:mark.edward.davis at gmail.com] On Behalf Of Mark Davis
Sent: Thursday, October 02, 2008 7:47 AM
To: Phillips, Addison
Cc: John Cowan; LTRU Working Group; Kent Karlsson

Subject: Re: [Ltru] Uniqueness of variant subtags

Mark

On Thu, Oct 2, 2008 at 4:29 PM, Phillips, Addison <addison at amazon.com> wrote:

I agree with John and I do not support Mark's proposal.

Mark: consider your own proposal to register 'hpin1958'. Would you be pleased if after you registered:

Type:variant
Subtag:hpin1958
Description: Hanyu Pinyin romanization of Mandarin Chinese

... somebody came along and added these descriptions to it:

Description: Pinyin romanization of Tibetan
Description: Tongyong romanization of Chinese
...

Or would that be a violation of "consistent semantics"? I think it is too hard to judge. Let's stick with semantic meaning.

As I said (but the threads got pretty convoluted), I could live with 'pinyin' being a broad term IF it were defined as a romanization specified by the Chinese government, and would be happy to change my registration form to reflect that. What I find bad is the various attempts to broaden pinyin without tying it to a particular organization.

Mark proposed this text:

<t>Records are unique by Type and Subtag: the registry MUST not

Ed.> MUST NOT

right.

contain two different records that have identical values for both Type
and Subtag. Thus requests to assign an additional record of a given
Type with an existing Subtag value MUST be rejected. For example, the
variant subtag 'rozaj' already exists in the registry, so adding a
second record of type 'variant' with the subtag 'rozaj' is prohibited.

AP> I liked my original version better, since it is more succinct.

I could live with your version.

However, information <i>can</i> be added to an existing record. For
example, the '1994' subtag variant record could have the prefix
'fr-CA' added together with a description of usage for a French
Canadian spelling reform associated with the year 1994.</t>

AP> I don't care for this proposal for the reasons given above.

I suppose we just disagree. Note that what I am writing is no more than giving an example of what is allowed by the current text and specification. We allow broadening of any term. And I would strongly disagree with any change to the current wording that would disallow that.

I don't know why you don't like this. If there were a 1994 spelling reform in France, I see nothing wrong with using fr-1994 - it is perfectly clear what that it means some version of french associated with the year 1994 - and if I want to find out more I look in the registry at the description. A darned sight easier to understand than most of our subtags!

For another thing, we can't let any particular language "get dibs" on a year - if we are going to use year numbers, then we have to be able to use them for different languages.

<t>Variant subtags that are used with multiple prefixes must have

Ed.> MUST have

good

consistent semantics across those prefixes. Variant subtags MAY

AP> "MAY" is too strong here. This text is more advisory than normative. I would suggest instead something like:

--
Requests for related variants might use a similar subtag format to visually indicate the relationship between them. For example, they might indicate an organization (such as a government or standards body). For example, 'ungegn' could be defined as referring to a transliteration for any given prefix as specified (etc..)

This wording doesn't really work for me, but I agree about removing the MAY. Let me try a rewording.

--

indicate an organization (including governments), such as 'ungegn',
'usbgn', or 'cisgost'. When applied to different prefixes, each would
consistently refer to a variant as specified by that organization for
those prefixes. For example, 'ungegn' could be defined as referring to
a transliteration for any given prefix as specified by the United
Nations Group of Experts on Geographical Names (UNGEGN).</t>

AP> Despite my proposal above, I think this is overkill. It suggests that some initial subtag sequences might acquire meaning that must be normatively enforced here and it doesn't solve a problem that is imminent.

I think it actually is imminent. Pinyin / wadegile are just the tip of an iceberg. We are working hard on romanizations, and there are dozens associated with given organization. I would find it abhorent to artificially distinguish subtags that are perfectly understandable as broad terms.

<t>Four digit subtags are reserved for indicating a year. Their
meaning is in reference to some significant specification or other
work associated with that year. It may have multiple prefixes: the
particular specification for a given prefix MUST be clearly indicated
in one of the Descriptions.</t>

AP> I think this is overkill. We don't have any examples of non-year four-digit registrations at present

That is why I want to spell out that they should always be used as years, if they are used.

nor do we have to spell out that a year should pertain to some event, I think. The registration process ought to do something useful, such as screen requests for such things. Besides, apparently the current trend is to append four letters of garbage to years :-).

No, it is to append four digits of garbage to words.

Addison

Addison Phillips
Globalization Architect -- Lab126

Internationalization is not a feature.
It is an architecture.

> -----Original Message-----
> From: ltru-bounces at ietf.org [mailto:ltru-bounces at ietf.org] On

> Behalf Of John Cowan
> Sent: Thursday, October 02, 2008 6:50 AM
> To: mark at macchiato.com
> Cc: LTRU Working Group; Kent Karlsson
> Subject: Re: [Ltru] Uniqueness of variant subtags
>

> mark at macchiato.com scripsit:
>
> > <t>Variant subtags that are used with multiple prefixes must have
> > consistent semantics across those prefixes.
>
> I'd say "a single meaning" rather than "consistent semantics".
> The hypothetical subtag 'western' has consistent semantics (it
> means
> "the western dialect of any language") but not a single meaning,
> for
> en-US-western has nothing to do with ko-western.
>
> --
> He made the Legislature meet at one-horse John Cowan
> tank-towns out in the alfalfa belt, so that cowan at ccil.org
> hardly nobody could get there and most of
> http://www.ccil.org/~cowan
> the leaders would stay home and let him go --H.L. Mencken's
> to work and do things as he pleased. Declaration of
> Independence

> _______________________________________________
> Ltru mailing list
> Ltru at ietf.org
> https://www.ietf.org/mailman/listinfo/ltru

Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.