Re: Suggested language for "mis" (Re: [Ltru] RE: ISO 639-2 decision:"mis")

We have used 'und' as a general "unknown" tag, which is really required for
effective operation of language tags. We are equating "und" with _no_tag_;
the latter are preferred in environments where that can be done, but "und"
is required where it is not allowed. And in a situation of inheritance (like
xml:lang), there needs to be something that says "don't inherit from the
parent level, because I'm not sure that I and my children are the same".

The key question for any tag -- assuming a tag is valid -- is what a
recipient can correctly infer from the presence of that tag. With "und" I
can make no assumptions about what the content is. It may be in no language,
some language, more than one language, a language that has no code. I can
simply make no inferences. As to why the original tagger tagged it that way,
I can't know -- maybe he knew didn't know what it was, maybe he didn't try
to find out, maybe he tried and didn't have 100% confidence, maybe he knew
it was low elvish and there wasn't a code for it, whatever.

Mark

On 6/17/07, Martin Duerst <duerst@it.aoyama.ac.jp> wrote:
>
> At 00:29 07/06/17, Mark Davis wrote:
> >That's going in the right direction, but doesn't not given enough
> guidance as to why to avoid it. My suggested language:
> >
> >The 'mis' (Uncoded) primary language subtag SHOULD NOT be used. According
> to ISO 639, it is used to identify linguistic content whose language is
> known but which does not *currently* have a corresponding subtag. It is thus
> intrinsically unstable -- the addition of other codes in the future can
> render its application invalid at any point without any warning -- and hence
> incompatible with the stability goals of BCP 47. It is thus always
> preferable to use other subtags: either "und" or -- with prior agreement --
> private use subtags.
>
> I'm a bit confused here. We want to try to give absoulte stability
> guarantees. But then we recommend 'und' (wrong to start with, because
> we know what language it is, so calling it undetermined is lying)
> or a private tag (inherently unstable and non-interoperable).
> Why don't we just accept that 'mis' is unstable (no disagreement
> here, of course), but instead of saying "invalid at any point without
> any warning -- and hence incompatible with the stability goals of BCP 47",
> warn the receiver that 'mis' may refer to a language that in the meantime
> has been coded, and change our language, e.g. 'invalid' -> 'suboptimal'.
>
> Regards,    Martin.
>
>
>
> #-#-#  Martin J. Du"rst, Assoc. Professor, Aoyama Gakuin University
> #-#-#  http://www.sw.it.aoyama.ac.jp       mailto:duerst@it.aoyama.ac.jp
>
>

-- 
Mark