Re: Xiangqi Game Symbols (was Re: Proposal to add standardized variation sequences for chess notation)

From: Michael Everson via Unicode <>
Date: Wed, 12 Apr 2017 12:54:58 +0100

On 12 Apr 2017, at 10:13, Andrew West via Unicode <> wrote:

> My Xiangqi proposal ( proposed a minimal set of logical game pieces for Xiangqi/Janggi, regardless of shape (circular or octagonal) or design (traditional characters, simplified characters, cursive characters, or pictures) which I consider a font design issue, and explicitly did not seek to encode circled ideographs. My proposal was rejected, and a different proposal by Michael Everson ( to encode all circled ideographs and negative circled ideographs attested in Xiangqi game diagrams was accepted instead.

Not quite. At the WG2 meeting it was proposed, I believe by experts from the US, to use circled ideographs to represent xiangqi characters. “In for a penny, in for a pound,” I thought, and so said that if we were to do that we’d have to encoded all the attested circled ideographs, because you can’t have a circled 士 (58EB) and say that a circled 仕 (4ED5) is a valid glyph variant of it. Then I wrote that proposal so that we could have an actionable document with which to get characters on the ballot.

> The accepted proposal for circled ideographs is a glyph encoding model not a character encoding model as for other game symbols (Chess,
> Dominos, Mahjong, Playing Cards, etc.),

This is true.

> and in my opinion it is a very bad model for several reasons. It makes the interchange of Xiangqi game data and game diagrams problematic; it hinders normal text processing operations on Xiangqi game pieces (for example, to search for a red horse piece you have to search for three different characters);

Yes, it does. It is important to remember that this use of symbols is a text usage.

> and in modern computer usage Xiangqi game pieces may not be represented as simple circled ideographs, but may be coloured designs showing characters or images.

Or black and white designs showing for instance an actual elephant rather than 象 8C61.

> It is also very likely that vendors will want to produce emoji versions of Xiangqi pieces,


> and these could not reasonably be considered to be glyph variants of circled ideographs.


> There has been some negative feedback on the circled ideographs model on the internet, and I believe that Michael has now been convinced that this model is wrong, and should be replaced by a model using logical game pieces.

I was convinced, and my proposal to rectify this were provided as Irish ballot comments to PDAM 1.2.

Michael Everson
Received on Wed Apr 12 2017 - 06:55:52 CDT

This archive was generated by hypermail 2.2.0 : Wed Apr 12 2017 - 06:55:53 CDT