On 12/12/2013 3:10 PM, Leo Broukhis wrote:
In the case of ɖ vs ð vs đ, there are three different letters, as
follows from their names, that happen to have identical capital glyphs
(those you've mentioned plus U+0110 LATIN CAPITAL LETTER D WITH STROKE).
Speaking of đ, "an alternate glyph with the stroke through the bowl is
used in Americanist orthographies" without any [loud] cries about
disunification.
If N-Eng and n-Eng are disunified but small engs aren't (should
they?), who keeps the "default" "toupper" conversion?
> And while they are at it, I wouldn't refuse if they squared the circle.
That's exactly right.
Leo
On Thu, Dec 12, 2013 at 2:52 PM, Asmus Freytag <[email protected]
<mailto:[email protected]>> wrote:
On 12/12/2013 2:25 PM, Leo Broukhis wrote:
Hmmm... As a person with Russian as the first language I can
assure you that from any literate Russian-speaking person's
perspective italic ū is an unacceptable and *WRONG*
representation of п (because in Russian, unlike Serbian, there is
й). Should we bother disunifying?
This example adds the issue of font style - because for styles
other than italic, the issue doesn't exist. I would take that as a
stronger indication that this is an issue that belongs in glyph
space.
The fact that the lowercase letter is the same in both cases
proves that the difference between N-Eng and n-Eng is purely
stylistic rather than semantic. Unicode shouldn't bother with
those minutia.
What about the reverse case, where the uppercase is the same and
the lower case isn't?
There are precedents in Unicode where these have been disunified.
U+00D0 LATIN CAPITAL LETTER ETH
U+0189 LATIN CAPITAL LETTER AFRICAN D
look exactly identical.
Precedents like this make the issue considerably less than clear cut,
> I suppose nothing will happen until the governments of
eng-using countries come together with a proposal.
Let's hope so. I wish they never do.
Lets hope they come together and endorse a solution that takes
into account not only rendering, but identifier security issues as
well. And while they are at it, I wouldn't refuse if they squared
the circle.
For some identifier systems it may be possible to institute a weak
normalization that would reduce the impact of duplicating the lower case
letter.
But disunifying any character at this stage may lead to either abandoned
data, or data that gets miscoded w/o anybody being the wiser.
Hence it is not true that
"Disunification is the best solution."
under every scenario.
A./
Leo
On Thu, Dec 12, 2013 at 2:06 PM, Michael Everson
<[email protected] <mailto:[email protected]>> wrote:
On 12 Dec 2013, at 15:29, Leo Broukhis <[email protected]
<mailto:[email protected]>> wrote:
> Hasn't
http://www.unicode.org/standard/where/#Variant_Shapes
explained it once and for all?
No, because users of N-shaped capital Eng consider n-shaped
capital Eng to be *WRONG*, not an acceptable variant. And
because n-shaped capital Eng consider N-shaped capital Eng to
be *WRONG*, not an acceptable variant.
Disunification is the best solution.
I suppose nothing will happen until the governments of
eng-using countries come together with a proposal.
Michael Everson * http://www.evertype.com/
