From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Fri Apr 01 2011 - 17:22:41 CST
On 4/1/2011 1:27 PM, Julian Bradfield wrote:
> On 2011-04-01, Michael Everson<everson@evertype.com> wrote:
>> Proposal to add heraldic hatching characters to the UCS
>> http://std.dkuug.dk/jtc1/sc2/wg2/docs/n4011.pdf
> (1) The supposed existing characters are characters encoded for
> round-trip compatibility with KS X 1001. They are not heraldic
> hatching characters, which don't exist and never have done.
> (2) Your proposal contains no evidence of plain text use. When I
> mentioned this elsewhere, you claimed that these characters were
> used in plain text - but as all your excerpts show, they
> aren't. The patterns are used in diagrams and tables, in the same
> way that any pattern can appear in a diagram or table, as for
> example on a map or on the legend of a map.
>
> Of course, the same goes for mah-jong, playing cards, and all sorts
> of other junk in Unicode. That's not a reason to add more junk. Every
> added character makes a lot of work for anybody trying to provide
> complete coverage, or complete non-Han coverage.
>
>
The characters that Michael proposes are clearly members of a set (with
rather well-defined shapes and relatively consistent semantics). That's
something that would need to be counted as points in favor. (Other
proposed characters had less to offer in that direction).
On the other hand, without more evidence of plain text use, there's a
good chance that these characters would very rarely see actual use, even
if encoded, and that documents would continue to use diagrams.
Finally,there are many other hatchings. The user interface of many
drawing programs gives 30+ different hatchings, so that the question
arises whether the heraldic subset is in fact the most useful one.
A./
PS: incidentally the source of the characters would not prevent them
being used for other semantics, but in this case, the annotation seems
to indicate the *absence* of any known general use. It is that absence
that is a stumbling block here. (Many symbols from KSC X 1001 are in
fact encoded as characters that are also used in other contexts).
This archive was generated by hypermail 2.1.5 : Fri Apr 01 2011 - 17:26:00 CST