From: Asmus Freytag (firstname.lastname@example.org)
Date: Fri Jan 02 2009 - 16:55:41 CST
On 1/1/2009 10:48 PM, Peter Constable wrote:
>> It is axiomatic that plain-text does not include little multicolored
>> pictures. Therefore, any application which inserts little multicolored
>> pictures in place of text strings is not a plain-text application.
> It's also axiomatic that the data passed through a plain-text protocol is plain text, a sequence of abstract characters, however that data may have originated at the source or be processed at its destination.
To elaborate further:
Where such abstract characters violate the character-glyph model for
Unicode, they are commonly known as "compatibility characters".
Arguments that certain characters fail various tests and criteria,
ultimately connected to the character-glyph model, have a different
relevance for compatibility characters than for other characters. That's
an important distinction that seems to not have fully sunk in, as
evidenced by the discussion so far.
Compatibility characters may need to be encoded for interoperability,
despite failing the normal tests for inclusion as characters. That has
been true in some way or other of all characters encoded for
compatibility so far. Depending on the nature of the compatibility, the
characters *may* obtain a compatibility decomposition, but that is not
universally so. The main criteria whether to include some character as
compatibility character are based on whether having it encoded is
necessary for interoperability with non-Unicode based plain text protocols.
>> Let's consider the committee as organism for a moment...
> I think you're going to great lengths while missing some simple realities that UTC sees before it:
> - data exchanged in plain-text protocols consists solely of abstract characters
> - the goal of the Universal Character Set is to be universal, implying (among other things) that any set of characters with significant usage in ICT industries must be considered potential candidates for encoding
> UTC has been and remains fully aware of the potential quagmires that can be encountered by starting to accept arbitrary kinds of graphic objects for encoding as characters, and is determined *not* to go in that direction. That, however, doesn't remove the aforementioned realities, and the way in which certain sectors of the ICT industry have thrust a *particular set* of graphic objects onto the world as abstract characters.
> It's as simple as that; no need for elaborate self-preservation conspiracy theories.
I fully agree with Peter here.
The only way to guard against future developments of this kind is to do
better outreach -- so that solutions can be found that don't evolve into
full blown, widely deployed non-Unicode based plain text protocols.
This archive was generated by hypermail 2.1.5 : Fri Jan 02 2009 - 18:28:30 CST