Norbert Bollow wrote:
> Andrew Winkler <email@example.com> wrote:
> > > > Formal and informal forms are, for the purposes of Majordomo, really
> > > > distinct languages. It will probably be easier and cleaner to treat
> > > > them as such. The only issue that remains is naming them.
> > >
> > > The more difficult issue is deciding what to serve when a user requests
> > > e.g. "German".
> > >
> > > -- NB.
> > How about an error message insisting on a more precise choice?
> Not a good idea. That would be counter-intuitive, and also it would not
> allow to honor the Prefer-Language: header (or any other similar internet
> standard which may emerge).
> -- NB.
If there is a real need for flavors of language ( and I agree that there
is ) then the Prefer-Language must reflect that, or it's broken. It's
reasonable, of course, to have the syntax hierarchical, so that
imprecise specifications default in some reasonable way. (Ie, no header
at all could be taken to mean "Chinese", German not otherwise modified
could be formal, in the least distinctive dialect, whatever that is;
for American English it's the dialect of Kansas; for Spanish that of
Bolivia and Peru, if my memory still serves. ) Anyway, presumably the
people working on the standard have bothered to inform themselves on the
range of linguistic issues.