Re: Control help- and Info-mode buffers from other buffers

[Manuel Giraud via "Emacs development discussions." <emacs-devel@gnu.org>]

Arthur Miller <arthur.miller@live.com> writes:

> Hi Manuel;
>
> thank you very much for testing it, and for the kind words. I have reworked
> somewhat the patch, because of some unnecessary warnings. I am attaching the
> reworked version in this mail too (I have send it in another mail). Sorry for
> the inconvenience, but if you test, I suggest testing with the later one. It is
> very much the same, minus some re-furnituring for the code, some local variables
> renaming and some doc string changes.

I'm testing this one but I think SPC should be 'help-scroll-up-command',
no?

[...]

> The technique can be extended to "remotely" control any buffer, but it works
> well only with "unique" buffers (one of a kind), like Help buffer for
> example. Even with *info* buffers it is a bit less clean since there can be few
> of those. But technically it is possible (I have been doing this
> programmatically too).

Yes, you're right: Man buffers do not have a unique name.  Maybe this
could be done later.

[...]

> I am not a repeat-mode user myself to be honest, so IDK, but I guess there
> should be no problems. These are just commands that can be called very well via
> M-x as any other command, so I guess it should work with repeat mode
> as well?

I, myself, have activated '(repeat-mode)' and I think your new bindings
are good candidate for it.  It could also come into another patch after
this one.

[...]

> But in Info-mode there are some nested char-tables in the keymap which turns out
> to be quite slow to process programmatically. Also, even if it was fast, doing
> that every time Emacs starts is unnecessary, it is probably better to do it once
> and for all, so I have implemented it now manually. Similar can be implemented
> for any command to make it "buffer specific", but as said, works well only in
> cases where there is just one buffer at a time, so I doubt in generality of this
> technique. I have actually tried to find some general way to do this
> programmatically with any buffer, but thus far, I don't see any general way
> that is both efficient and 100% failsafe.

Ok, I think I start to see your motivations for this patch then.  And I
do think that having this in place for *Help* buffers and *info* makes
sense.
-- 
Manuel Giraud