registry vs. extensions...

Addison Phillips [wM] aphillips at
Mon Oct 20 15:19:30 CEST 2003

John Cowan replied:
> > First, I think it might be worthwhile for there to be a policy of not
> > registering new language (or region) based tags until the
> relevant "base"
> > standard (ISO639 or ISO3166) has rejected a request for registration.
> The whole point of RFC 3066 registration is that it's lightweight and
> easy compared to ISO registration.  This is no longer as true as it
> was in the past, but it's still true.

My point is that registering tags that then become part of ISO639 is kludgey
and confusing. And that there is a tendency for folks not to go to ISO639
first and least take that registration authority's temperature (so to
speak). If a particular standard isn't cutting the mustard, having another
standard to work around it temporarily is not necessarily the right

I agree that it is easier to register with the RFC than with ISO, but speed
isn't necessarily the only consideration. These tags are essentially
forever. So why not strive to get things right? If we had a duck test to
figure out whether ISO were inclined (or not) to register a particular value
we could save a lot of top-level registrations that quickly become outmoded.
> > I think it might be time to consider an alternative.
> > Specifically, rfc3066 needs an extension mechanism that *doesn't* use
registration. Sort of a
> > "private use area" for language tags.
> Any tag that begins with "x-" is private use.
Again, not my point. An "x-" tag means "the tag that follows is my private
tag". You lose the utility of any registered or standardized subtags
contained therein. There is no way to say "de-DE-x-mySubtag". In fact, the
tag registry and x- prefixes are sort of all-or-nothing affairs. The text of
the RFC implies that a registered tag can almost be considered a single
unitary value, possibly unrelated to any other values and not at all
permitted to use the generative mechanism.

If I read it correctly, a search for "x-de-DE-mySubtag" does NOT find
"de-DE". "de-DE at mySubtag" would find such a match. It would be a way of
extending the generative mechanism for use with items that do not rise to
the level of general usage. There is still the option of registering a tag
(and there are obviously cases where that is required over the use of an
extension), but it might reduce the need to create fully-realized, complex,
registered sets of tags and solve some of the dialect/minority language
issues grappled with on this list in the past.

Again, just an idea (and not even one that is original to me).



More information about the Ietf-languages mailing list