CE Whitehead cewcathar at
Wed Sep 3 18:57:28 CEST 2008

Hi!  I think Niall Tracey's made a good point here--that the prefix needs to be [zh-latn] so that there is fault tolerance where applications do not recognize the subtag [wadegile].  (They recognize few variant subtags as far as I can tell.)


--C. E. Whitehead
cewcathar at

From: "Tracey, Niall" niall.tracey at

> We all know that new tags don't gain sudden acceptance
> overnight. It would be one thing if there was a suppress-script
> for subtags -- in that case systems could update themselves
> automatically. As it is, that script information is in the free-text
> fields and requires human intervention to insert into systems, so
> it's going to be a while coming. In legacy systems, it may never
> happen.
> If we issue advice saying that the text should be tagged zh-
> Latn-wadegile, it will be rendered correctly by all properly-
> written software systems.
> If we tell people that zh-wadegile is enough, more of them will
> tag text as such and systems will attempt to render it in Hanji.

Agreed; applications do not immediately recognize new subtags such as [wadegile].

> Izh-Latn-wadegile is surely the only fault-tolerant (userproof?)> option.
> Shouldn't we build fault-tolerance in at every level?

> N?all.

More information about the Ietf-languages mailing list