RE: RE: Proposing new arrow characters with Bidi_Mirrored=Yes

Nitai Sasson via Unicode <[email protected]> Thu, 10 Apr 2025 22:37:14 +0000
Newsgroups gmane.text.unicode.general
Message-ID <[email protected]>
On Friday, 11 April 2025 at 01:18, Doug Ewell <[email protected]> wrote:

> But the tricky part, as you stated in your first message, is that in some cases arrows need to follow the visual directionality of surrounding characters, while in other cases they need to ignore surrounding characters and truly mean “left” or “right.”
> 
> This is why adding the existing arrows to the “mirrored” list would merely swap one problem for another, and why you suggested a mirroring control character.
> 
> --
> Doug Ewell, CC, ALB | Lakewood, CO, US | ewellic.org

Exactly. I never seem to choose my words well enough - as you correctly understood, when I said "make available", I meant with the use of a control character. This behavior _cannot_ be applied to existing unmodified arrow characters - it would violate the Unicode stability policy.

My original suggestion was to add new arrow characters which would each have the Bidi_Mirrored=Yes proprety. MANY new arrow characters. Every existing horizontal arrow would have needed a variant with this property. This was not the best idea, but it started the conversation.

The current suggestion is to apply this behavior to existing characters using a modifier/control character.

Thank you for helping to clarify the current status.

I have some thoughts about the ramifications of such a control character, and other control characters that naturally "fall out" when considering it. I feel like I have convinced enough of the participants here that this idea has merit, so I will share these thoughts here soon, under a new title.