Encoding scripts in tags: evil or just unpleasant?

[Peter_Constable at sil.org]

John Cowan wrote on 05/23/2003 06:16:26 AM:

> > >Well, I started out in your position and have now moved to approving
of
> > >Peter's productive langage-script-country model (where "country" is a
> > >proxy for spelling system, basically).
> >
> > Edberg I suppose you mean...

> Sorry for (obviously) worse confusing the issue; I was referring to Peter
> _Constable_'s model as expressed in "Language identification and IT"

Four years ago I was similarly reluctant when the idea of incorporating ISO
15924 into RFC1766bis (now RFC3066) was being considered. After mulling it
over for three years, I came to believe the problems could be understood
well enough to do it. Discussions that have gone on on this list and some
offlist (which haven't endorsed everything I suggested in that paper -- not
that I *ever* expected it was all worth keeping) have helped me see there
are still some things to work through (specifically how it would relate to
the whole "locale" stuff), but it has also convinced me all the more that
we're very close to ready to take this step.

And as I have said before, yi-latn wasn't the first step in this direction
as far as registrations are concerned; it should be completely obvious to
all that de-1996 et al were the first registrations that went in this
direction.




- Peter


---------------------------------------------------------------------------
Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485