el-latn, ru-latn, and related possibilities

Tex Texin tex at xencraft.com
Fri Oct 7 08:02:35 CEST 2005


John,

"John.Cowan" wrote:
> 
> Tex Texin scripsit:
> 
> > It seems wasteful however to specify both transliteration and script, since
> > script is implied by the transliteration scheme. Size of tags is somewhat a
> > concern for my users. But it probably isn't a compelling argument for this
> > list.
> 
> It's wasteful only if you assume that every recipient of a language tag
> knows all the transliteration codes and what scripts they imply.
> That's not likely.

Perhaps true, but easily solved by publishing the info, perhaps along with
the registration of the translit code.

> > transliteration is more important than region/dialect
> 
> Maybe, maybe not; it depends on the language.  Greeks, at least those
> young enough to use computers, can typically read four or five
> el-Latn-* systems with ease.

True. Of course, some transliteration schemes are used across many
languages.
When I buy language books, I notice many of them make the same choices for
representing pronunciation.

So, if we coded translit-* my reader might work for several languages, which
could be fine for some applications (e.g. learning songs to sing.) If I were
more choosy I could select translit-el...
I am not arguing translit should come first, just questioning the
significance of one or two use cases.
I agree though for people, you are right, we manage to deal with a number of
variations of translit schemes pretty well.
I think software processing the text needs more information though.
 
> > zh-Latn-pinyin-cn is better than zh-Latn-cn-pinyin.
> 
> Probably zh-Latn-pinyin is the sensible tag.

Sure, unless we want to distinguish TW, HK, SG, CN, et al. as we do for
zh-Hans, and zh_Hant...

John, I am not trying to be argumentative. I am looking for compelling
arguments, rather than the one or two use cases (unless the use cases are so
clearly the typical or dominant scenario).
I can agree that translit isn't always necessary, just as region isn't
always needed.
If it is needed, does it make sense for it to be a variant after region or
is it better ahead of region?
Peter's comment seemed to me to support it being ahead of region. I don't
see here a counter argument either, only the consideration that it isn't
always needed.

tex


More information about the Ietf-languages mailing list