Re: Repeat undo-only is not working

[Jim Porter <jporterbugs@gmail.com>]

On 8/23/2021 11:40 PM, Juri Linkov wrote:
>>> Sorry, this part of the feature is still unpolished.
>>> The intention was to disallow repeating of undo with
>>> the key sequence `C-/ u', but allow only with `C-x u u'.
>>> Maybe this was a wrong idea?  Do you think it should be
>>> possible to type `C-/' to initiate the repeating sequence
>>> `C-/ u u u'?
>>
>> For what it's worth, the intention matches my personal expectation. [snip]
>>
>> Perhaps it should be possible to support the behavior in the original post,
>> but I'm happy with how things are now.
> 
> So there is a need to distinguish between these cases
> 
>    C-/ --- u u u
>    C-x u --- u u
> 
> ("---" visually separates the initial and repeating sequence)
> and allow their customization.  Maybe a new variable is sufficient
> that will inhibit checking that the last character exists in repeat-map.
> 
> But what to do if the user wants to disable C-/ --- u u u,
> but still wants to use for gdb-step such sequence
> C-x C-a C-n --- n n n n where repeat-map has no C-n?

If there were a flag like `repeat-enable-aggressively' to enable things 
like C-/ --- u u u, then would it be possible to let a user who doesn't 
want C-/ --- u u u but *does* want C-x C-a C-n --- n n n to set that 
flag to nil and then call some function in their .emacs for each "extra" 
sequence in the repeat-map they want to enable? Another option would be 
to set `repeat-enable-aggressively' to t and then *remove* any repeat 
mappings the user doesn't like.

I haven't looked at the implementation, so I'm not sure if either of 
those options are feasible, but they'd give users plenty of flexibility 
if they really wanted it.

That said, I'd probably be happy with setting 
`repeat-enable-aggressively' to nil in my config and not worrying about 
overriding any specific cases.

- Jim