From: verdy_p (verdy_p@wanadoo.fr)
Date: Sun Feb 01 2009 - 21:59:02 CST
De : "Peter Constable"
> I'm not sure what "still supported by the ISO 639 authorities" means.
I think he is refering to the maintenance and registration authorities (MA's and RA's) for each of the parts of the
ISO 639 standards suite.
What is not clear is the meaning of "still supported".
For me there's still some support for the old versions of the standard, at least for bookkeeping and preserving
interoperability for long with legacy applications that were conforming and can't be decided to become magically
non-conforming. That's why the standards are versioned (and there's a bookkeeping of the dated changes).
Note however that ISO 639 did not have any major release until quite recently (less than 3 years ago). The changes
that occured since its first publication were considered "minor" and were not erratas. They had no mandatory status
in the standard itself before the approval and publication of the major release. We are still not far enough from
this change to say that the codes that have been "deprecated" since "long" are in fact completely "deleted" and
then forbidden in confoming applications.
There are LOTS of applications (and many more users of these applications) that have not been reviewed since this
major release (just look at Windows: there are still users of Windows 98 despite it is 12 years old; and even it is
not sold and can't be bought with the support from Microsoft, it is still used in many systems, sometimes in
embedded form with a specific licence; 3 years ago, XP was still stable and it is still widely used, but it was not
designed to support the new major version of ISO 639).
There's a difference between "supporting" and "recommending". In fact NO standard can be mandatory, everything is a
recommendation at the date of publication and reflects the state of the current consensus.
Even if Unicode was created a long time ago, it is not "old" and subject to deprecation. In fact Unicode has just
started this year to become predominant (with its UTF-8 form) to all other legacy 8-bit or multibyte encodings on
the Web, but the decline of other 8-bit encodings (notably ASCII, ISO 3166-1 and Shift-JIS and their minor variants
in Windows codepages) is now slowing: we can still say that Unicode is still very new for many users (about half of
total users on the web, this still means billions people, even if there are now as many people that are aware of it
and use it since more or less long!)
There is simply too many documents to convert, and documents that won't be converted at all but will remain
published "as is", as long as there are people interested in them and someone paying for their hosting or
archiving. But there are still new documents created today that use the legacy encodingsn, because they are created
on older systems that are still not capable of using the full Unicode capabilities, and there's no money for these
users to upgrade their systems or no time to invest (this is money too) for converting the documents they need and
use since long.
My opinion is that "old" can only apply to technical standards that have NOT been in use by a majority of people
since more than 10 years (We are still very far from this date with Unicode, you'll have to wait for another
decennial, possibly more or indefinitely for ASCII): this is very different from the effectively short lifecycle of
commercial applications (as long as they are supported by their initial promoter or copyright holder); there's much
life long after the "official support", and even sometimes long after the end of exclusive rights on intelletual
properties (e.g. the "GIF" format now fully in the public domain but still widely used: even if there's nobody
supporting it "officially", it is is still supported by lot of people and used by legions.)
This archive was generated by hypermail 2.1.5 : Sun Feb 01 2009 - 22:02:02 CST