From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Tue Mar 10 2009 - 18:29:38 CST
On 3/10/2009 4:19 PM, Michael Everson wrote:
> On 10 Mar 2009, at 22:57, Kenneth Whistler wrote:
>
>> BTW, this discussion is recapitulating positions taken
>> a dozen years ago, back in late 1996 and early 1997, when
>> we were debating the addition of what became U+20AC EURO SIGN.
>> Some folks back then wanted to just change the glyph
>> of the existing U+20A0 EURO-CURRENCY SIGN -- and we had
>> to push up FAQs for awhile after Unicode 2.1 came out
>> emphasizing that the EURO SIGN was U+20AC and *NOT*
>> U+20A0. Fortunately, the right decisions were taken back
>> then, and things eventually sorted themselves out correctly.
>
> Um, No, because the EURO-CURRENCY SIGN actually pointed to a
> *different* currency than the EURO SIGN did.
No, slight correction: there was *speculation* as to what it might point
to. But it isn't clear that it ever pointed to an *actual* currency.
The rupee being used in different countries, it would be an exceedingly
bad idea to modify the glyph.
The character code doesn't encode the function, it encodes the graphic
symbol. When that changes to a different graphic symbol, you need a new
character.
There's an archiving aspect to this that you seem to not be aware of. In
the digital age, documents exist that contain just a font style and a
plaint text character code. When these documents are reproduced years
later, you do not want their appearance to be magically altered, just
because a bad decision in the character coding community forced all
fonts to be updated.
In other words, if I used Times in 1991 to create a document with Rs,
then I want to be able to load that in 2019 and show it with the old
style for the Rs symbol, not whatever new symbol was invented, because,
in 1991 that new symbol didn't exist.
This is different from your favorite type of glyph correction: that
case, where the standard inadvertently mixes font designs within a table
of character codes, because sample fonts came from different sources. In
such cases, the glyph change doesn't signify a different graphical
symbol, but just tries to emphasize the fact that each font should use a
*consistent* design across all its glyphs for that set of characters.
For letters, and common punctuation, the typography is often settled
enough that font vendors need little guidance from the Unicode code
tables - for symbols the question of what are allowable/required
variations is a lot less settled, so great care must be taken to not
inadvertently raise questions or doubt about the identity of the
character in question.
A./
>
>
>
This archive was generated by hypermail 2.1.5 : Tue Mar 10 2009 - 18:31:25 CST