From: Werner LEMBERG (wl@gnu.org)
Date: Sat Mar 15 2008 - 23:51:37 CST
> Your image is an artificial example, just reading "foo bar baz", not
> a convincing document that exhibits specific textual semantics or
> need for correct orthography.
This is correct. I'm not a philologist, and I my main interest is
basically to assign correct character codes to two glyphs in a Greek
encoding which has been used for that purpose.
> The two tips at ends look also similar to ends that may also be used
> in underlined groups. If you had multicolored inks or wanted to
> display on screen you could as well exhibit this emphasizing using
> background colors; you could as well use a surrounding box.
Well, I think that this nexus is based on existing documents, however,
I don't know them. My basic hope was that people on this list know
this emphasizing construction.
> May be something with semantic value could be encoded in text, but
> it would be something like grouping parentheses with various
> alternatives and no required presentation.
Yes, I'm searching grouping parentheses suitable for this.
> So may be you could use the existing parenthese pairs (notably the
> existing tortoise shell parentheses that are the nearest from your
> exhibited image) [...]
I've mentioned in my mail that this is exactly what I want to use,
however, there isn't any sample usage given un UTR 25.
> Note that existing parenthese symbols are not naturally grouping by
> themselves: [...]
The mathematical brackets as mentioned in UTR 25 do group.
> Are you proposing to encode the horizontal half-parentheses and let
> the text renderer decide if it will support the connection line
> between each half or if it will make this connection controled by
> simple underlining/overlining styles?
This is the idea, yes, to mark the start and end of the nexus (similar
to horizontal brackets or braces), and the text rendering engine does
the right formatting. Certainly, such a thing has already be encoded
in Unicode, I believe, at least for mathematical purposes.
Werner
This archive was generated by hypermail 2.1.5 : Sun Mar 16 2008 - 00:26:20 CST