Re: Proposal: Remap interactive commands via keymaps

[Kim F. Storm]

Jason Rumney <address@hidden> writes:

> address@hidden (Kim F. Storm) writes:
> 
> > Why should C-k and M-x kill-line be treated differently ?
> 
> Because if a user goes to the trouble of M-x kill-line, then they
> really want kill-line;


Maybe, but often users have to go to the trouble of M-x command
because there is no explicit binding for that command (or they just
don't remember where the command is, but do remember what it's called,
so they use M-x command).  And in those cases, the command *must* be
remapped!

>     if they wanted whizz-bang-replacement-kill-line,
> they would use C-k, or M-x whizz-bang-replacement-kill-line.

But what if I an in a mode where C-k doesn't run kill-line, so
I have to enter M-x kill-line to run it....   Then I don't want to
have to remember to run M-x whizz-bang-replacement-kill-line to run
the proper(!) version of kill-line when whizz-bang-mode is active!

C-k vs. M-x kill-line may be a poor example here, but I really don't
see why they should be treated differently.  If I enable my-minor-mode
which modifies the behaviour of kill-line to my-kill-line, then C-k
and M-x kill-line should run my-kill-line.

> 
> There should always be a way for the user to use the original command.

Hm...  If you use defadvice on kill-line, how do you run the original
kill-line command?

But in any case, if the user really wants to confuse things, he should
be able to do so.  But why make this the default?

Maybe we could make `ESC M-x command' run the original `command' if
that is really what the user wants (but that will definitely be an
exception - and thus for expert users anyway).

-- 
Kim F. Storm <address@hidden> http://www.cua.dk