Re: Should U+3248 ... U+324F be wide characters?

[Philippe Verdy via Unicode]

But today's new emojis do not come all from Japanese. Many new emojis zre
not even tied to Asian cultures. I don't see why a candle, or train, or F1
racing car would necessarily be square, they would look just ugly (with
excessive horizontal margins, or smaller than needed with excessive
vertical margins).
Same thing about new emopjis showing groups of people (e.g. a family with
two adults and two children).
For emojis showing a single person face (emoticons) there's however no
problem to fit them in a square, same things about the various symbols in
squares or circles used in Japanese.

2017-08-18 23:29 GMT+02:00 Peter Edberg :

> Per UTS #51 (see http://www.unicode.org/reports/tr51/#Design_Guidelines):
>
> "Current practice is for emoji to have a square aspect ratio, deriving
> from their origin in Japanese. For interoperability, it is recommended that
> this practice be continued with current and future emoji. They will
> typically have about the same vertical placement and advance width as CJK
> ideographs.'
>
> - Peter E
>
> On Aug 18, 2017, at 1:48 PM, Philippe Verdy via Unicode <
> unicode@unicode.org> wrote:
>
> I don't think that emojis are necessarily "square", they could be larger
> (e.g. a train or a snake or an horizontal railway, or a group of several
> peoples, or a cloud) or narrower (e.g. a candle).
>
> Rendering them as square will make sense only in contexts where this makes
> sense ** if possible** : monospaced fonts. But there are cases where a
> single character cell would not be enough and multiple cells would be
> needed (notably in text terminals, but as well in sinographic contexts
> uwing multiple em-squares in a row).
>
> The classification of widths in CJK if there to help determine how many
> cells will be needed in two cases: narrow rectangular cells used in text
> terminals, or square cells in classic sinographic typesetting (which is
> still not mandatory because variable-width rendering is also possible, even
> if it is less common, using more specific fonts for such artistic use or to
> correctly render handwritten calligraphy). This classification of widths
> makes no sense in Latin where it variable-width is still prefered and more
> common.
>
> So there will be both variants for variable-width and "monospaced"
> (cell-based) rendering of emojis, like they both exist for CJK and Latin:
> Latin letters has a "narrow" width in sinographic square contexts only to
> allow two letters side-by-side per square instead of centering them with
> wide gaps or rendering them in widdened variants. Most Asian emojis from
> CJK charsets will render in a single square cell, but others may still need
> two square cells for better rendering (without having to use variable width
> that would break the grid layout).
>
> When rendering Latin words in CJK contexts, the alignment to the grid may
> also be made only on spans of Latin letters (one or more words), by
> recentering it in a row of as many cells that could fit: it would be even
> more useful for Arabic sequences. This technic however would not fit very
> well in classic "text terminals" where half-width Latin, Hebrew and Arabic
> will still be preferable (or full-width for some Arabic letters with some
> extenders, or some long Arabic ligatures).
>
>
>
> 2017-08-18 14:21 GMT+02:00 Andre Schappo :
>
>>
>> On 18 Aug 2017, at 00:50, Philippe Verdy via Unicode 
>> wrote:
>>
>>
>> 2017-08-17 18:46 GMT+02:00 Asmus Freytag (c) via Unicode <
>> unicode@unicode.org>:
>>
>>> On 8/17/2017 7:47 AM, Philippe Verdy wrote:
>>>
>>> 2017-08-17 16:24 GMT+02:00 Mike FABIAN via Unicode 
>>> :
>>>
 Asmus Freytag via Unicode  さんはかきました:
 Most emoji now have "W", for example:

 1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS

 That seems correct because emoji behave more like Ideographs.

 Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
 This seems to me also more like an Ideograph.
>>>
>>>
>>> Not really. They have existed since extremely long without being bound
>>> to ideographs or sinographic requirements on metrics. Notably their
>>> baseline and vertical extension do not follow the sinographic em-square
>>> layout convention (except when they are rendered with CJK fonts, or were
>>> encoded in documents with legacy CJK encodings, also rendered with suitable
>>> CJK fonts being then prefered to Latin fonts which won't use the large
>>> siongraphic metrics).
>>>
>>> If they were like emojis, they would actually be larger : I think it is
>>> a case for definining a Emoji-variant for them (where they could also be
>>> colored or have some 3D-like look)
>>>
>>>
>>> There's an emoji variant for the standard digits.
>>>
>>
>> Do you speak about circled numbers ? I don't think so.
>>
>> I (and Mike as well to which I was replying) was speaking about a good
>> case for defining emoji 

Re: Should U+3248 ... U+324F be wide characters?

[Peter Edberg via Unicode]

Per UTS #51 (see http://www.unicode.org/reports/tr51/#Design_Guidelines 
):

"Current practice is for emoji to have a square aspect ratio, deriving from 
their origin in Japanese. For interoperability, it is recommended that this 
practice be continued with current and future emoji. They will typically have 
about the same vertical placement and advance width as CJK ideographs.'

- Peter E

> On Aug 18, 2017, at 1:48 PM, Philippe Verdy via Unicode  
> wrote:
> 
> I don't think that emojis are necessarily "square", they could be larger 
> (e.g. a train or a snake or an horizontal railway, or a group of several 
> peoples, or a cloud) or narrower (e.g. a candle).
> 
> Rendering them as square will make sense only in contexts where this makes 
> sense ** if possible** : monospaced fonts. But there are cases where a single 
> character cell would not be enough and multiple cells would be needed 
> (notably in text terminals, but as well in sinographic contexts uwing 
> multiple em-squares in a row).
> 
> The classification of widths in CJK if there to help determine how many cells 
> will be needed in two cases: narrow rectangular cells used in text terminals, 
> or square cells in classic sinographic typesetting (which is still not 
> mandatory because variable-width rendering is also possible, even if it is 
> less common, using more specific fonts for such artistic use or to correctly 
> render handwritten calligraphy). This classification of widths makes no sense 
> in Latin where it variable-width is still prefered and more common.
> 
> So there will be both variants for variable-width and "monospaced" 
> (cell-based) rendering of emojis, like they both exist for CJK and Latin: 
> Latin letters has a "narrow" width in sinographic square contexts only to 
> allow two letters side-by-side per square instead of centering them with wide 
> gaps or rendering them in widdened variants. Most Asian emojis from CJK 
> charsets will render in a single square cell, but others may still need two 
> square cells for better rendering (without having to use variable width that 
> would break the grid layout).
> 
> When rendering Latin words in CJK contexts, the alignment to the grid may 
> also be made only on spans of Latin letters (one or more words), by 
> recentering it in a row of as many cells that could fit: it would be even 
> more useful for Arabic sequences. This technic however would not fit very 
> well in classic "text terminals" where half-width Latin, Hebrew and Arabic 
> will still be preferable (or full-width for some Arabic letters with some 
> extenders, or some long Arabic ligatures).
> 
> 
> 
> 2017-08-18 14:21 GMT+02:00 Andre Schappo  >:
> 
>> On 18 Aug 2017, at 00:50, Philippe Verdy via Unicode > > wrote:
>> 
>> 
>> 2017-08-17 18:46 GMT+02:00 Asmus Freytag (c) via Unicode 
>> >:
>> On 8/17/2017 7:47 AM, Philippe Verdy wrote:
>>> 2017-08-17 16:24 GMT+02:00 Mike FABIAN via Unicode >> >:
>>> Asmus Freytag via Unicode >> > さんはかきました:
>>> Most emoji now have "W", for example:
>>> 
>>> 1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS
>>> 
>>> That seems correct because emoji behave more like Ideographs.
>>> 
>>> Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
>>> This seems to me also more like an Ideograph.
>>>  
>>> Not really. They have existed since extremely long without being bound to 
>>> ideographs or sinographic requirements on metrics. Notably their baseline 
>>> and vertical extension do not follow the sinographic em-square layout 
>>> convention (except when they are rendered with CJK fonts, or were encoded 
>>> in documents with legacy CJK encodings, also rendered with suitable CJK 
>>> fonts being then prefered to Latin fonts which won't use the large 
>>> siongraphic metrics).
>>> 
>>> If they were like emojis, they would actually be larger : I think it is a 
>>> case for definining a Emoji-variant for them (where they could also be 
>>> colored or have some 3D-like look)
>> 
>> There's an emoji variant for the standard digits.
>> 
>> Do you speak about circled numbers ? I don't think so.
>> 
>> I (and Mike as well to which I was replying) was speaking about a good case 
>> for defining emoji variant of these circled (or squared) numbers (Mike spoke 
>> about circled number 10, which is not encoded as an emoji and not even as an 
>> ideograph, and that he proposed to give a wide width property like 
>> ideographs).
>> 
>> 
> 
> Are not CJK ideographs both (W)ide and (S)quare? Does (W)ide imply or define 
> that the ideograph should also be (S)quare?
> 
> It seems to me that there are many characters that are both (W)ide and 
> 

Re: Should U+3248 ... U+324F be wide characters?

[Philippe Verdy via Unicode]

I don't think that emojis are necessarily "square", they could be larger
(e.g. a train or a snake or an horizontal railway, or a group of several
peoples, or a cloud) or narrower (e.g. a candle).

Rendering them as square will make sense only in contexts where this makes
sense ** if possible** : monospaced fonts. But there are cases where a
single character cell would not be enough and multiple cells would be
needed (notably in text terminals, but as well in sinographic contexts
uwing multiple em-squares in a row).

The classification of widths in CJK if there to help determine how many
cells will be needed in two cases: narrow rectangular cells used in text
terminals, or square cells in classic sinographic typesetting (which is
still not mandatory because variable-width rendering is also possible, even
if it is less common, using more specific fonts for such artistic use or to
correctly render handwritten calligraphy). This classification of widths
makes no sense in Latin where it variable-width is still prefered and more
common.

So there will be both variants for variable-width and "monospaced"
(cell-based) rendering of emojis, like they both exist for CJK and Latin:
Latin letters has a "narrow" width in sinographic square contexts only to
allow two letters side-by-side per square instead of centering them with
wide gaps or rendering them in widdened variants. Most Asian emojis from
CJK charsets will render in a single square cell, but others may still need
two square cells for better rendering (without having to use variable width
that would break the grid layout).

When rendering Latin words in CJK contexts, the alignment to the grid may
also be made only on spans of Latin letters (one or more words), by
recentering it in a row of as many cells that could fit: it would be even
more useful for Arabic sequences. This technic however would not fit very
well in classic "text terminals" where half-width Latin, Hebrew and Arabic
will still be preferable (or full-width for some Arabic letters with some
extenders, or some long Arabic ligatures).



2017-08-18 14:21 GMT+02:00 Andre Schappo :

>
> On 18 Aug 2017, at 00:50, Philippe Verdy via Unicode 
> wrote:
>
>
> 2017-08-17 18:46 GMT+02:00 Asmus Freytag (c) via Unicode <
> unicode@unicode.org>:
>
>> On 8/17/2017 7:47 AM, Philippe Verdy wrote:
>>
>> 2017-08-17 16:24 GMT+02:00 Mike FABIAN via Unicode :
>>
>>> Asmus Freytag via Unicode  さんはかきました:
>>> Most emoji now have "W", for example:
>>>
>>> 1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS
>>>
>>> That seems correct because emoji behave more like Ideographs.
>>>
>>> Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
>>> This seems to me also more like an Ideograph.
>>
>>
>> Not really. They have existed since extremely long without being bound to
>> ideographs or sinographic requirements on metrics. Notably their baseline
>> and vertical extension do not follow the sinographic em-square layout
>> convention (except when they are rendered with CJK fonts, or were encoded
>> in documents with legacy CJK encodings, also rendered with suitable CJK
>> fonts being then prefered to Latin fonts which won't use the large
>> siongraphic metrics).
>>
>> If they were like emojis, they would actually be larger : I think it is a
>> case for definining a Emoji-variant for them (where they could also be
>> colored or have some 3D-like look)
>>
>>
>> There's an emoji variant for the standard digits.
>>
>
> Do you speak about circled numbers ? I don't think so.
>
> I (and Mike as well to which I was replying) was speaking about a good
> case for defining emoji variant of these circled (or squared) numbers (Mike
> spoke about circled number 10, which is not encoded as an emoji and not
> even as an ideograph, and that he proposed to give a wide width property
> like ideographs).
>
>
>
> Are not CJK ideographs both (W)ide and (S)quare? Does (W)ide imply or
> define that the ideograph should also be (S)quare?
>
> It seems to me that there are many characters that are both (W)ide and
> (S)quare eg emoji
>
> André Schappo
>
>

Re: Should U+3248 ... U+324F be wide characters?

[Andre Schappo via Unicode]

On 18 Aug 2017, at 00:50, Philippe Verdy via Unicode 
> wrote:


2017-08-17 18:46 GMT+02:00 Asmus Freytag (c) via Unicode 
>:
On 8/17/2017 7:47 AM, Philippe Verdy wrote:
2017-08-17 16:24 GMT+02:00 Mike FABIAN via Unicode 
>:
Asmus Freytag via Unicode > 
さんはかきました:
Most emoji now have "W", for example:

1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS

That seems correct because emoji behave more like Ideographs.

Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
This seems to me also more like an Ideograph.

Not really. They have existed since extremely long without being bound to 
ideographs or sinographic requirements on metrics. Notably their baseline and 
vertical extension do not follow the sinographic em-square layout convention 
(except when they are rendered with CJK fonts, or were encoded in documents 
with legacy CJK encodings, also rendered with suitable CJK fonts being then 
prefered to Latin fonts which won't use the large siongraphic metrics).

If they were like emojis, they would actually be larger : I think it is a case 
for definining a Emoji-variant for them (where they could also be colored or 
have some 3D-like look)

There's an emoji variant for the standard digits.

Do you speak about circled numbers ? I don't think so.

I (and Mike as well to which I was replying) was speaking about a good case for 
defining emoji variant of these circled (or squared) numbers (Mike spoke about 
circled number 10, which is not encoded as an emoji and not even as an 
ideograph, and that he proposed to give a wide width property like ideographs).



Are not CJK ideographs both (W)ide and (S)quare? Does (W)ide imply or define 
that the ideograph should also be (S)quare?

It seems to me that there are many characters that are both (W)ide and (S)quare 
eg emoji

André Schappo

Re: Should U+3248 ... U+324F be wide characters?

[Mike FABIAN via Unicode]

"Asmus Freytag (c) via Unicode"  さんはかきました:

> On 8/17/2017 7:24 AM, Mike FABIAN wrote:
>> Asmus Freytag via Unicode  さんはかきました:
>>
>>> On 8/16/2017 6:26 AM, Mike FABIAN via Unicode wrote:
>>>
>>>  EastAsianWidth.txt contains:
>>>   3248..324F;A # No [8] CIRCLED NUMBER TEN ON BLACK
>>> SQUARE..CIRCLED NUMBER EIGHTY ON BLACK SQUARE
>>>   i.e. it classifies the width of the characters at
>>> codepoints
>>>  between 3248 and 324F as ambiguous.
>>>   Is this really correct? Shouldn’t they be “W”, i.e. wide?
>>>   In most fonts these characters seem to be square shaped
>>> wide characters.
>>>
>>> "W" not only implies display width, but also a different treatment in the 
>>> context of line
>>> breaking and vertical layout of text.
>>>
>>> "W" characters behave more like Ideographs, for the most part, while "N" 
>>> are treated as
>>> forming words (for the most part).
>> Most emoji now have "W", for example:
>>
>> 1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS
>>
>> That seems correct because emoji behave more like Ideographs.
>>
>> Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
>> This seems to me also more like an Ideograph.
>>
>>> "A" means, you get to decide whether to treat these as "W" or "N" based on 
>>> context. If
>>> used in a non ideographic context, they behave like all other symbols (but 
>>> happen to fill
>>> an EM square).
>
> "A" means, you get to decide whether to treat these as "W" or "N" based on 
> context.
>
> There's really not strong need to change an "A" towards "W", because
> "A" doesn't get in your way if you decided that "W" works better for
> you.
>
> Remember that all the EAW properties ares supposed to be "resolved"
> down to W or N. For some, like Na that resolution is deterministic,
> for A it is context/application dependent, but when you finally
> process your data, only W(ide) or N(arrow) remain after resolution.

OK, that means that is OK to decide that in the context of glibc
resolving these to W(ide) is best, right?

-- 
Mike FABIAN 

Re: Should U+3248 ... U+324F be wide characters?

[Philippe Verdy via Unicode]

2017-08-17 18:46 GMT+02:00 Asmus Freytag (c) via Unicode <
unicode@unicode.org>:

> On 8/17/2017 7:47 AM, Philippe Verdy wrote:
>
> 2017-08-17 16:24 GMT+02:00 Mike FABIAN via Unicode :
>
>> Asmus Freytag via Unicode  さんはかきました:
>> Most emoji now have "W", for example:
>>
>> 1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS
>>
>> That seems correct because emoji behave more like Ideographs.
>>
>> Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
>> This seems to me also more like an Ideograph.
>
>
> Not really. They have existed since extremely long without being bound to
> ideographs or sinographic requirements on metrics. Notably their baseline
> and vertical extension do not follow the sinographic em-square layout
> convention (except when they are rendered with CJK fonts, or were encoded
> in documents with legacy CJK encodings, also rendered with suitable CJK
> fonts being then prefered to Latin fonts which won't use the large
> siongraphic metrics).
>
> If they were like emojis, they would actually be larger : I think it is a
> case for definining a Emoji-variant for them (where they could also be
> colored or have some 3D-like look)
>
>
> There's an emoji variant for the standard digits.
>

Do you speak about circled numbers ? I don't think so.

I (and Mike as well to which I was replying) was speaking about a good case
for defining emoji variant of these circled (or squared) numbers (Mike
spoke about circled number 10, which is not encoded as an emoji and not
even as an ideograph, and that he proposed to give a wide width property
like ideographs).

Re: Should U+3248 ... U+324F be wide characters?

[Asmus Freytag (c) via Unicode]

On 8/17/2017 7:24 AM, Mike FABIAN wrote:

Asmus Freytag via Unicode  さんはかきました:


On 8/16/2017 6:26 AM, Mike FABIAN via Unicode wrote:

 EastAsianWidth.txt contains:
 
 3248..324F;A # No [8] CIRCLED NUMBER TEN ON BLACK SQUARE..CIRCLED NUMBER EIGHTY ON BLACK SQUARE
 
 i.e. it classifies the width of the characters at codepoints

 between 3248 and 324F as ambiguous.
 
 Is this really correct? Shouldn’t they be “W”, i.e. wide?
 
 In most fonts these characters seem to be square shaped wide characters.


"W" not only implies display width, but also a different treatment in the 
context of line
breaking and vertical layout of text.

"W" characters behave more like Ideographs, for the most part, while "N" are 
treated as
forming words (for the most part).

Most emoji now have "W", for example:

1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS

That seems correct because emoji behave more like Ideographs.

Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
This seems to me also more like an Ideograph.


"A" means, you get to decide whether to treat these as "W" or "N" based on 
context. If
used in a non ideographic context, they behave like all other symbols (but 
happen to fill
an EM square).


"A" means, you get to decide whether to treat these as "W" or "N" based on 
context.

There's really not strong need to change an "A" towards "W", because "A" doesn't get in 
your way if you decided that "W" works better for you.

Remember that all the EAW properties ares supposed to be "resolved" down to W 
or N. For some, like Na that resolution is deterministic, for A it is context/application 
dependent, but when you finally process your data, only W(ide) or N(arrow) remain after 
resolution.

A./



A./

Re: Should U+3248 ... U+324F be wide characters?

[Asmus Freytag (c) via Unicode]

On 8/17/2017 7:47 AM, Philippe Verdy wrote:



2017-08-17 16:24 GMT+02:00 Mike FABIAN via Unicode 
>:


Asmus Freytag via Unicode > さんはかきました:
Most emoji now have "W", for example:

1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS

That seems correct because emoji behave more like Ideographs.

Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
This seems to me also more like an Ideograph.

Not really. They have existed since extremely long without being bound 
to ideographs or sinographic requirements on metrics. Notably their 
baseline and vertical extension do not follow the sinographic 
em-square layout convention (except when they are rendered with CJK 
fonts, or were encoded in documents with legacy CJK encodings, also 
rendered with suitable CJK fonts being then prefered to Latin fonts 
which won't use the large siongraphic metrics).


If they were like emojis, they would actually be larger : I think it 
is a case for definining a Emoji-variant for them (where they could 
also be colored or have some 3D-like look)


There's an emoji variant for the standard digits.

A./

Re: Should U+3248 ... U+324F be wide characters?

[Philippe Verdy via Unicode]

2017-08-17 16:24 GMT+02:00 Mike FABIAN via Unicode :

> Asmus Freytag via Unicode  さんはかきました:
> Most emoji now have "W", for example:
>
> 1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS
>
> That seems correct because emoji behave more like Ideographs.
>
> Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
> This seems to me also more like an Ideograph.


Not really. They have existed since extremely long without being bound to
ideographs or sinographic requirements on metrics. Notably their baseline
and vertical extension do not follow the sinographic em-square layout
convention (except when they are rendered with CJK fonts, or were encoded
in documents with legacy CJK encodings, also rendered with suitable CJK
fonts being then prefered to Latin fonts which won't use the large
siongraphic metrics).

If they were like emojis, they would actually be larger : I think it is a
case for definining a Emoji-variant for them (where they could also be
colored or have some 3D-like look)

Re: Should U+3248 ... U+324F be wide characters?

[Mike FABIAN via Unicode]

Asmus Freytag via Unicode  さんはかきました:

> On 8/16/2017 6:26 AM, Mike FABIAN via Unicode wrote:
>
> EastAsianWidth.txt contains:
> 
> 3248..324F;A # No [8] CIRCLED NUMBER TEN ON BLACK SQUARE..CIRCLED 
> NUMBER EIGHTY ON BLACK SQUARE
> 
> i.e. it classifies the width of the characters at codepoints
> between 3248 and 324F as ambiguous.
> 
> Is this really correct? Shouldn’t they be “W”, i.e. wide?
> 
> In most fonts these characters seem to be square shaped wide characters.
>
> "W" not only implies display width, but also a different treatment in the 
> context of line
> breaking and vertical layout of text.
>
> "W" characters behave more like Ideographs, for the most part, while "N" are 
> treated as
> forming words (for the most part).

Most emoji now have "W", for example:

1F600..1F64F;W   # So[80] GRINNING FACE..PERSON WITH FOLDED HANDS

That seems correct because emoji behave more like Ideographs.

Isn’t this the same for “CIRCLED NUMBER TEN ON BLACK SQUARE”?
This seems to me also more like an Ideograph.

> "A" means, you get to decide whether to treat these as "W" or "N" based on 
> context. If
> used in a non ideographic context, they behave like all other symbols (but 
> happen to fill
> an EM square).
>
> A./
>

-- 
Mike FABIAN 
睡眠不足はいい仕事の敵だ。

Re: Should U+3248 ... U+324F be wide characters?

[Asmus Freytag via Unicode]

  
  
On 8/16/2017 7:23 AM, Philippe Verdy
  via Unicode wrote:

These
  squared (or circled) characters however do not have registered
  variants for digits with descenders (used in traditional
  typographic fonts for Latin), such as 4, 7 or 9, or variable-width
  digits (not using the more modern digits with "figure-space"
  fixed width), but I think the later would not require such variant
  given that it's more the width of the enclosing square (or circle)
  which is important, and digits will be adjusted in width and
  interdigit gaps, as needed to fit.
This ^^ is a bit of a non-sequitur. From an
encoding perspective, we don't care what types of digits a font
uses for these; users will want the total with of the symbols to
be fixed so they work well in numbering lists, etc., but that
doesn't constrain the interior design. I'm not surprised that
there are no registered variants, because some of the number
styles, like descenders, really wouldn't look good. 
  
End of digression.
A./
  
  

Re: Should U+3248 ... U+324F be wide characters?

[Philippe Verdy via Unicode]

I do agree, only CJK fonts using in CJK contexts will render them as "W"
(i.e. the fixed-width srandard ideogaphic composition square). If they are
used in Latin, they will adopt the metrics of the Latin font including
them, thery will be square but not necessarily aligned with the ideographic
square but could be aligned to that their internal digits will align on the
same base as normal digits, and the sqaure will include the Latin descender
and ascender height, the width will be adapted to match it (digits may need
to be compacted horizontally to fit the square with those metrics, but will
preserve their baseline alignment. If the normal Latin digits don't have
descenders, the squared variants may not include the full height used by
Latin letters with descenders.

These squared (or circled) characters however do not have registered
variants for digits with descenders (used in traditional typographic fonts
for Latin), such as 4, 7 or 9, or variable-width digits (not using the more
modern digits with "figure-space" fixed width), but I think the later would
not require such variant given that it's more the width of the enclosing
square (or circle) which is important, and digits will be adjusted in width
and interdigit gaps, as needed to fit.

2017-08-16 16:04 GMT+02:00 Asmus Freytag via Unicode :

> On 8/16/2017 6:26 AM, Mike FABIAN via Unicode wrote:
>
> EastAsianWidth.txt contains:
>
> 3248..324F;A # No [8] CIRCLED NUMBER TEN ON BLACK SQUARE..CIRCLED 
> NUMBER EIGHTY ON BLACK SQUARE
>
> i.e. it classifies the width of the characters at codepoints
> between 3248 and 324F as ambiguous.
>
> Is this really correct? Shouldn’t they be “W”, i.e. wide?
>
> In most fonts these characters seem to be square shaped wide characters.
>
>
> "W" not only implies display width, but also a different treatment in the
> context of line breaking and vertical layout of text.
>
> "W" characters behave more like Ideographs, for the most part, while "N"
> are treated as forming words (for the most part).
>
> "A" means, you get to decide whether to treat these as "W" or "N" based on
> context. If used in a non ideographic context, they behave like all other
> symbols (but happen to fill an EM square).
>
> A./
>

Re: Should U+3248 ... U+324F be wide characters?

[Asmus Freytag via Unicode]

  
  
On 8/16/2017 6:26 AM, Mike FABIAN via
  Unicode wrote:


  EastAsianWidth.txt contains:

3248..324F;A # No [8] CIRCLED NUMBER TEN ON BLACK SQUARE..CIRCLED NUMBER EIGHTY ON BLACK SQUARE

i.e. it classifies the width of the characters at codepoints
between 3248 and 324F as ambiguous.

Is this really correct? Shouldn’t they be “W”, i.e. wide?

In most fonts these characters seem to be square shaped wide characters.



"W" not only implies display width, but also
a different treatment in the context of line breaking and
vertical layout of text.
"W" characters behave more like Ideographs,
for the most part, while "N" are treated as forming words (for
the most part).
"A" means, you get to decide whether to
treat these as "W" or "N" based on context. If used in a non
ideographic context, they behave like all other symbols (but
happen to fill an EM square).
A./