[tex-hyphen] New language classiclatin – question about the language code

[Mojca Miklavec]

Dear Jonathan,

On Thu, Jun 5, 2014 at 2:42 PM, Jonathan Kew wrote:
> On 4/6/14 14:10, Mojca Miklavec wrote:
>
>> No, I didn't say it wasn't acceptable. I just mentioned that once
>> la-classic is registered, el-classic may not be registered later for
>> example.
>
> I'm not sure this is correct. You wouldn't register "la-classic" as such;
> you'd register "classic" as a variant subtag. And as far as I understand
> (see http://tools.ietf.org/html/bcp47#section-3.5), you could register the
> variant "classic" for use with multiple prefixes, so that both "la-classic"
> and "el-classic" would be legitimate.
>
> The important point would be to make it clear in the initial registration
> that the "classic" variant subtag is intended to mark "classical"
> orthography for any language that has a classical vs modern distinction, and
> NOT to describe it as a tag specifically for "classical Latin".
>
> The "prefix" field in the registry could either be omitted altogether,
> leaving the variant free to be used with any language (like "fonipa"), or
> multiple prefixes could be listed: at least "la" and "el", but there are no
> doubt other reasonable candidates as well (classical Sanskrit, Tamil, ...).
> Compare, for example, the registry entry for "baku1926".
>
> Note that it is legal to add further prefixes in subsequent requests,

I probably took the following sentence too seriously:

    Requests to add a 'Prefix' field to a variant subtag that imply a
    different semantic meaning SHOULD be rejected

Indeed it should be ok if the prefix is used to imply "the same
semantic meaning".

> so it
> is not necessary to be exhaustive in the initial registration. But if
> including a prefix at all, it is, I think, important to make it clear in the
> description, and by including several example prefixes, that the usage of
> this variant subtag is NOT intended to be limited to a specific language.

Thanks a lot. That clarifies a few things. I feel better about this
after reading your explanation.

Mojca

PS: About "baku1926": I find it useful that the tag contains something
in addition to the year (other examples include "1606nict",
"1694acad", "luna1918" ...). If another language standardized
something in 1926 or 1918, they are free to pick another letter
combination together with the year and then both subtags could
co-exist. The situation with "1994" and "1996" is a bit absurd in that
respect. They could have registered "1994res" or "1996ger" for that
matter, but I understand that this was probably one of the early
additions when "de-1996" was registered/meant as a complete tag.

Of course it's debatable whether "Classical Latin" and "Classical
Tamil" are semantically the same thing, but it should be acceptable
for this purpose, I believe.