Re: [Geiser-users] Overriding the M-[backquote] key binding

[Jose A. Ortega Ruiz]

On Thu, Jan 13 2011, Mark Harig wrote:

>>
>> >> I wanted to provide an alternative to C-. because that key combo
>> >> is not recognized in some terminals (notably, the one i use,
>> >> urxvt), and M-` is nicely close to M-TAB in US keyboards. I
>> >> kind of like it, but i guess i should be nice and leave the
>> >> standard binding alone. Any suggestions for an alternative?
>> >>
>> >
>> > Some possible candidates:
>> >
>> >   C-c ` (backquote)
>> >   C-c TAB
>> >   M-S-TAB (M-BACKTAB)
>> >   C-TAB
>>
>> These are good candidates (i would probably favour C-c TAB), but now
>> that i've thought of it, i think the right key binding is C-u TAB,
>> because geiser-completion--complete-module is a special case of
>> completion-at-point (the latter completes on identifiers and module
>> names, and both do completion).
>>
>> What do you think?
>>
>
> You might want to look at the Emacs function `complete-symbol', which
> accepts a prefix argument.  I would like for Geiser's behavior to be
> as consistent as it can be made to be with existing Emacs behavior so
> that people are not surprised when using it and so that they don't
> have to learn two different behaviors for the same key bindings.  If
> you anticipate that `geiser-repl--tab' capabilities might one day more
> closely match the behavior of `complete-symbol', then don't bind C-u
> TAB to `geiser-completion--complete-module'.

No, i dpn't think so. They actually already differ substantially from
complete-symbol. In the first place, completion is done by
completion-at-point (which has no C-u variant); moreover,
geiser-repl--tab will indent or go to the next error (instead of
completing), depending on context. Finally, the alternative behaviour of
complete-symbol is performing a lookup in the manual's index, which we
have already covered (sort of) by other Geiser commands and, IMHO, was a
horrible choice for a C-u binding (it being a substantially different
function). If we deemed it useful to provide an "index lookup" command
(and here there's the problem of non-texinfo manuals), i would strongly
prefer to give it an entirely different key binding, unrelated to
completion (it'd belong to the C-c C-d family).

Cheers,
jao
-- 
You err by thinking simplicity and elegance are mostly
cosmetic.  Simplicity and elegance are overwhelmingly
practical virtues.
  - William D Clinger, comp.lang.scheme