leob at mailcom.com
Mon Jul 6 10:53:27 CDT 2015
On Mon, Jul 6, 2015 at 8:18 AM, Doug Ewell <doug at ewellic.org> wrote:
> Leo Broukhis <leob at mailcom dot com> wrote:
>> Most platforms display unknown printable characters as white
>> rectangles with hex digits in them.
>> In Doug's message, I saw a rectangle with 01F in the upper row, and
>> 3F3 in the lower row.
> This is a handy feature, at least for character geeks like us, but "most
> platforms" might be a bit misleading here. There is a rather commonly
> used platform that starts with the letter W which does not do this.
I was a little surprised myself when I saw it in Firefox under W7
Enterprise, but here we are.
>> Moreover, on any platform when users see unknown characters, they
>> search for a font, install it and are able to see in cleartext at
>> least something they can make sense of. For a RIS or any other
>> non-default-ignorable character on a non-vexillology-aware platform, a
>> font with stylized letters would be sufficient to read the intent of
>> the writer, and, as a free extra, to tell apart Liechtenstein and
>> Haiti without squinting.
> I think a useful bit of feedback on PRI #299 would be to inquire whether
> it is, in fact, a design goal to handle this use case of transparency of
Huh? What kind of a deliberate design goal would be to forgo semantics
in favor of presentation, even as a fallback behavior?
In an ideal world, where all platforms are actively maintained, and
all maintainers rush to implement the cool new features,
it could have been acceptable, but not in our world, I'm afraid.
> the individual letters on platforms, rendering engines, and/or fonts
> that don't support flag-tag composition. (Please, not
> "non-vexillology-aware." None of these platforms studies or analyzes
> flags. They assemble multiple characters into a single image.)
"Vexillology awareness" was, of course, mostly in jest.
> If transparency on flag-tag-unaware platforms is not a design goal, it
> might be difficult to make the case that default-ignorable tag
> characters are a poor choice because they don't support transparency.
Right. Then the objection should be interpreted with regard to the design goal.
More information about the Unicode