From: Mark Davis (mark.edward.davis@gmail.com)
Date: Wed May 27 2009 - 17:26:06 CDT
> Fine.
>
> Whatever.
>
> No need to try to apply standards.
Again, no need to be snide.
Both Peter and I wouldn't be doing what we do if we didn't believe in the
need for standards. Had you actually followed some of the links I posted in
this thread, you would have seen the specification for the transform IDs.
I'll explain briefly, since you didn't follow the link.
The IDs have a source and a target and optionally a variant, in the form
source-target/variant. The source and target can each be Unicode language
IDs, Unicode script IDs (long or short), or other ids. The Unicode
language/script IDs are basically BCP 47 but with some small
extensions/restrictions and the use of "_" instead of "-" in the canonical
form (the syntactic differences are for backwards compatibility). So an
example is:
ru_RU-en_US/BGN
While we could use fonipa, a valid source or target would then be a language
tag, like "en-fonipa", so we would end up with a transform ID like
"en-en_fonipa". However, especially since we are intending to use this as a
pivot between scripts, the fallback behavior of "en_fonipa" isn't ideal, and
it was simpler to just use a custom value. If instead of "fonipa" we
actually had a real script ID for IPA (eg 'Fipa'), then that would work. But
we don't have the equivalent of 'Fipa', so a custom value works better (an
alternative would be to use a PU script code).
Mark
On Wed, May 27, 2009 at 07:41, Michael Everson <everson@evertype.com> wrote:
> On 27 May 2009, at 15:33, Peter Constable wrote:
>
> It's en_US-fonipa, not en_US-ipa
>>>
>>
>> ??!!
>>
>> The use of an underscore would make this clearly NOT a BCP 47 tag. If it's
>> BCP 47, it would be en-US-fonipa; if it's not a BCP 47 tag, it can be
>> whatever the usage context specifies, including perhaps en_US-ipa.
>>
>
> Oh.
>
> Fine.
>
> Whatever.
>
> No need to try to apply standards.
>
> I'm wrong.
>
>
> Michael Everson * http://www.evertype.com/
>
>
>
This archive was generated by hypermail 2.1.5 : Wed May 27 2009 - 17:29:13 CDT