Unicode input regression

No response under 
<https://lists.freebsd.org/archives/freebsd-current/2023-September/004521.html> 
or 
<https://mail.kde.org/pipermail/kde-freebsd/2023-September/040457.html>, 
so I'm casting the net wider.

Can freebsd-desktop subscribers help?

Whilst input failure is not a showstopper, it /is/ a significant 
inconvenience.

Thanks

-------- Forwarded Message --------
Subject: 	Unicode input regression
Date: 	Fri, 1 Sep 2023 17:57:56 +0100
From: 	Graham Perrin <grahamperrin@gmail.com>
To: 	FreeBSD CURRENT <freebsd-current@freebsd.org>, kde@freebsd.org


Following my first pkg upgrade with 15.0-CURRENT, I no longer get the 
required character after input.

In applications such as Dolphin, Konsole, Firefox, and Thunderbird: the 
visible input (e.g. code 2013 for an en dash) is followed by disappearance.

In Code - OSS (editors/vscode): input is followed by visibility of the 
code (e.g. 2013), as if I typed it normally.

The version of IBus did not change.

Does anyone else have this regression?