Re: C-c / r key-setting bug (?)

["Andreas Röhler" <andreas.roehler@easy-emacs.de>]

Am 21.10.2010 16:04, schrieb Carsten Dominik:
>
> On Oct 21, 2010, at 3:34 PM, Andreas Röhler wrote:
>
>> Am 21.10.2010 15:00, schrieb Carsten Dominik:
>>>
>>> On Oct 21, 2010, at 2:42 PM, Andreas Röhler wrote:
>>>
>>>> [ ... ]
>>>>>>> C-c / will prompt for another key and then dispatch depending on
>>>>>>> that
>>>>>>> key.
>>>>>>> Many do.
>>>>>>
>>>>>> Hmm, that would fit into the already mentioned coding-style category.
>>>>>> Is it wise to do it that way?
>>>>>> For example M-x describe-mode fails telling about these keys.
>>>>>> Are reasons for this?
>>>>>
>>>>> Yes. There are not enough keys, unless I introduce a larger number
>>>>> of full-scale prefix keys. And it is much easier to remember just `C-c
>>>>> /' and then get a friendly prompt for a number of options. For
>>>>> interactive use, I think this is just perfect.
>>>>>
>>>>> Now, if you want to write programs based on these functions, then you
>>>>> want to have the real command names. This is one of the reasons why I
>>>>> like to have them in the manual, for look up.
>>>>>
>>>>
>>>> Hi,
>>>>
>>>> I'm following, thanks for the explanation.
>>>> Remains a describe-mode bug than (?)
>>>
>>> I would not call it a bug. Describe-mode looks at the key maps, it
>>> cannot look into the functions.
>>>
>>
>> OK, but can't you make it known to the keymap?
>
> No, I cannot.
>
> - Carsten
>
>
>
>

Ah, we can, I'm sure.
Let's assume you assigned the same keys to different commands.
Somehow Emacs must discriminate the states where the one or the other 
are to use.
So let's call these states modes, make a keymap for it, no?

Andreas