suppress-script values for fil, mi, pes, prs, qu members

John Cowan cowan at ccil.org
Thu Oct 21 23:05:38 CEST 2010


On Thu, Oct 21, 2010 at 10:40 AM, Peter Constable
<petercon at microsoft.com> wrote:

> At this point, John Cowan has raised a concern, but he
> hasn't rebutted my last reply to him.

I hadn't previously considered the use of S-S in extensions to 4647
tag matching: you receive a document tagged foo-Cyrl-ZZ, and you see
that the S-S for foo is Cyrl, so you change it to just foo-ZZ before
matching.  That is very plausible.

I withdraw my concern.

However, I register another concern: is evidence at hand for all the
Quechua languages that they are in fact written?  It is inappropriate
to provide S-S data for unwritten languages. (Mere transcription does
not count.)

Doug suggested that we should have inferred from an S-S record in a
macrolanguage that the individual records should have the same S-S
value, but that is unsafe, not only for the reasons given in this
thread, but also because macrolanguage groupings are mutable: a
language may be added to the group that uses a different script.


More information about the Ietf-languages mailing list