From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Fri Oct 26 2007 - 03:03:42 CDT
On 10/25/2007 10:41 PM, vunzndi@vfemail.net wrote:
> Even more could have been done at the start of unicode and CJKV.
>
> An even more effcient solution as far as code points, would have been
> to encode ...
In theory, and with lots of hindsight, perhaps.
But in the political and technical reality of the early years of Unicode
(and ISO10646), there was not a whole lot of room for radically
different approaches.
For a new technology to be successful, you need to have both enough new
features to make adoption of the technology worth the effort, and a
migration strategy (strategy of small steps) that allows porting to the
new technology in manageable and relatively predictable ways.
There are many design elements in Unicode that were adopted for
migration purposes - some were less necessary in hindsight than others;
likewise some of them proved more costly than expected, even to
implementations not requiring those particular migrations.
One of the benefits of hindsight is that the true magnitude and scope of
certain collections has become much better understood as these
collections have become cataloged in Unicode. Also, the deficiencies of
particular cataloging choices are now well-known and understood.
As a result, alternative representations, building on the amassed
knowledge base, but distinct in their approach, could be created, that
may prove potentially superior at certain tasks. Whether such
alternative representations would be limited to rendering, processing,
data compression or whether they would be attractive as interchange
format remains very much to be seen (if and when somebody sits down and
actually cranks out such a system).
But whatever form it takes, it would be subject to the same kind of
forces that shaped Unicode's early stages, except that it's 20 years
later, and general state of the art and the knowledge of scripts,
characters, their use and implementation have not stood still. So, while
the influence of these forces may well lead to different particular
results now, it does not mean that very different approaches would have
been feasible then.
A./
>
>
>
>
This archive was generated by hypermail 2.1.5 : Fri Oct 26 2007 - 03:06:35 CDT