Randy Williams asked:
> Folks,
>
> This is not directly a Unicode question, but I hope people
> don't mind my posting this here.
>
> I hope you can settle a debate I am having with a co-worker.
>
> Given a font that is ship on Windows where this font has various
> glyphs replaced in the 0x80 - 0x9F range with 22 "Box Drawing"
> characters. These are the 22 "Box Drawing" characters that are
> found in cp850 on IBM PC and are found in U+2500 plane of Unicode.
> These 22 "Box Drawing" characters do not exist in Windows cp1252
> encoding. So the replacement is done to add those characters to the
> font and the font otherwise is exactly like cp1252.
>
> One of us claims that this font has effectively created a new
> encoding. The other claims that this is just a new font. Which
> is it?
>
Well, actually, you are both right. I hope you didn't have
money riding on the outcome.
This is an example of the practice, widespread on the Macintosh,
of "font-encoded text". Someone creates an arbitrary font and
assigns arbitrary glyphs to existing code points, without
changing the script system. This is how, for example, Klingon
gets "encoded" on the Mac.
But font-encoded text doesn't really change the character encoding.
It simply creates a private convention usable only by those who
have access to that particular font. It won't be recognized by
standards--even vendor standards, with few exceptions--nor be
processed by other than custom convertors. As a result, most
font-encoded text is untransmissible. Archives of font-encoded
text gradually become uninterpretable as the people familiar with
the particular font conventions they used go on to other things.
People should switch to Unicode-based solutions wherever they
can meet their needs instead, since Unicode-encoded text will
remain interpretable for an indefinite time into the future.
--Ken
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:36 EDT