[discuss] Possible approaches to solving "problem no. 1"

Andrew Sullivan ajs at anvilwalrusden.com
Fri Feb 21 01:35:33 UTC 2014

On Fri, Feb 21, 2014 at 01:24:51AM +0100, Michel Gauthier wrote:

> globalization? How do you understand the Unicode "globalization" and
> the normative internationalization.

What is "normative" internationali[s|z]ation?

I think it will be important in this discussion to distinguish among
at least three terms.  One of them is "globalization", the meaning of
which seems in this thread to be up in the air.  Therefore, I'll not
talk about it.

A second is "internationalization" or "internationalisation" or "i18n"
(for "i-18-letters-n"), which in the Internet protocol world is really
just about preparing parts of a system for use by diverse language
users.  "Just use Unicode" is an i18n slogan.  The point in this case
is that there are things that are exposed to diverse user communities
and they need to be available to them in ways that are convenient for
them.  This is what (for instance) IDNA is about: making a protocol
_available_ to support users.

A third is "localization" or "localisation" or "l10n".  In the
Internet protocol (or, more accurately, "above" the Internet
protocols) world, _this_ is the thing that is intended for users.
I18n is useless for users.  Nobody wants a user interface
simultaneously in English and Farsi and Chinese.  L10n is intended to
make the user-facing features sensitive to the user's local
conditions, to the extent that is automatically determinable.  We have
learned a great deal about this issue over the past 15 or 20 years,
but there is more yet to do.  If I may be permitted a modest plug, the
IAB's Internationalization program (we use USian spelling in the IAB
these days) is in fact currently engaged in tackling an update to RFC
2277.  It's early days, however.

I hope these distinctions are helpful to the current discussion.

Best regards,


Andrew Sullivan
ajs at anvilwalrusden.com

More information about the discuss mailing list