Philip Kaludercic <[email protected]> writes:

> This is related to bug#60101.
>
> We could fix this upstream by checking for a symbol property, say
> `canonical-binding', and prefer that if FIRSTONLY is non-nil.

Thanks for the insight and for pointing me to bug#60101.

The proposed change sounds good.  It would simplify the code here and
remove the keymap-copying workaround.  However, I would prefer not to
make this patch depend on that change.  As I mentioned, I first
considered implementing this over two years ago and have kept postponing
it since then.  We can use the workaround for now and simplify the
implementation once the upstream support lands.

Best,
-- 
Slawomir Grochowski

Reply via email to