Encoding italic (was: A last missing link)

[Victor Gaultney via Unicode]

Andrew Cunningham wrote:
> Underlying, bold text, interletter spacing, colour change, font style 
> change all are used to apply meaning in various ways. Not sure why 
> italic is special in this sense.

Italic is uniquely different from these in that the meaning has been 
well-established in our writing system for centuries, and is 
consistently applied. The only one close to being this is use of 
interletter spacing for distinction, particularly in the German and 
Czech tradition. Of course, a model that can encode span-like features 
such as italic could then support other types of distinction. However 
the meaning of that distinction within the writing system must be clear. 
IOW people do use colour change to add meaning, but the meaning is not 
consistent, and so preserving it in plain text is relatively pointless. 
Even Bold doesn't have a consistent meaning other than <strong> - but 
that's a separate conversation.

> And I am curious on your thoughts, if we distinguish italic in 
> Unicode, encode some way of spacifying italic text, wouldn't it make 
> more sense to do away with italic fonts all together? and just roll 
> the italic glyphs into the regular font?

That's actually being done now. OpenType variation fonts allow a variety 
of styles within a single 'font', although I personally feel using that 
for italic is misguided.

The reality is that the most commonly used Latin fonts - OS core ones - 
all have italic counterparts, so app creators only have to switch to 
using that counterpart for that span. And if the font has no italic 
counterpart then a fallback mechanism can kick in - just like is done 
when a font doesn't have a glyph to represent a character.

> In theory changing italic from a stylistic choice as it currently is 
> to a encoding/character level semantic is a paradigmn shift.

Yes it would be - but it could be a beneficial shift, and one that more 
completely reflects distinctions in the Latin script that go back over 
400 years.

> But it it were introduced I would prefer a system that was more 
> inclusive of all scripts, giving proper analysis of typeseting and 
> typographic conventions in each script and well founded decisions on 
> which should be encoded. Cherry picking one feature relevant to a 
> small set of scripts seems to be a problematic path.

The core issue here is not really italic in Latin - that's only one 
case. An adjusted text model that supports span-like text features, 
could also unlock benefits for other scripts that have consistent 
span-like features.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://unicode.org/pipermail/unicode/attachments/20190117/0d6cbb15/attachment.html>