On 08/02/2024 18:12, Dmitry Gutov wrote:
> On 08/02/2024 16:43, Dmitry Gutov wrote:
>> If you can come up with a code that detects at run time that help-key
>> and/or F1 was rebound to a key that will not invoke
>> describe-prefix-bindings, such a key should indeed better be removed
>> from the message.  But can we reliably do that?  If we cannot, having
>> two keys there instead of one is better.
> 
> Here's a rough draft.
> 
> It seems to work in the basic cases that I've tried (changing help-char 
> to something with a binding and rebinding <f1>), but doesn't account for 
> key translations for far (e.g. if help-char is ?X and the prefix map has 
> ?x, this isn't caught).
> 
> Also, piping the current used map through so many methods is pretty 
> messy, I'm sure whether I've used the appropriate value in other 
> callsites of echo_now and echo_dash. There's also echo_update...
> 
> So if anybody has something simpler in mind that'd be welcome.

Here's that simpler version.

It doesn't address the "key translations" example above, but it seems 
rare enough, and it should be possible to fix later.

So I suggest we install this now.