<tab> differs from TAB, tab-key differs from C-i. What?!

<tab> differs from TAB, tab-key differs from C-i. What?!

[cons]

Hello! I'm very confused about one thing:

Go to *scratch*-buffer and press 'C-h b' (describe-bindings). Under
'Major Mode Bindings' TAB is bound to lisp-indent-line. This mean that
when I press the tab-key or C-i (the same thing), lisp-indent-line is
executed. Fine.

BUT, now try this. Evaluate these 2 expressions in *scratch*:

(defun foo ()
  (interactive)
  (message "Executed foo."))
(global-set-key (kbd "<tab>") 'foo)

Once again, press 'C-h b'. Our new binding of <tab> to foo is found
under 'Global Bindings'. So, we have the old local binding (major
mode) of TAB to lisp-indent-line that shadows a global binding of
<tab> to foo. Now pressing the tab-key or C-i should mean the same
thing (produce ascii code 9), which in turn should execute lisp-indent-
line (since local bindings takes precedence over global bindings),
right?

Wrong. Pressing the tab-key executes foo, and pressing C-i executes
lisp-indent-line. So, TAB seems to mean pressing C-i, which apparently
differs from <tab> which seems to mean pressing the tab-key. Now, this
is becoming a mess.

1) Shouldn't pressing the tab-key and C-i mean the same thing?
2) Shouldn't all emacs' representations of a tab (<tab>, TAB, ?\t,
"\t", ascii code 9, <C-i>, ?\C-i, ?\^i etc etc) mean the same thing?

Best regards,

/Stefan

Re: <tab> differs from TAB, tab-key differs from C-i. What?!

[cons]

On 22 Aug, 12:26, cons <conss...@gmail.com> wrote:
> Hello! I'm very confused about one thing:
>
> Go to *scratch*-buffer and press 'C-h b' (describe-bindings). Under
> 'Major Mode Bindings' TAB is bound to lisp-indent-line. This mean that
> when I press the tab-key or C-i (the same thing), lisp-indent-line is
> executed. Fine.
>
> BUT, now try this. Evaluate these 2 expressions in *scratch*:
>
> (defun foo ()
>   (interactive)
>   (message "Executed foo."))
> (global-set-key (kbd "<tab>") 'foo)
>
> Once again, press 'C-h b'. Our new binding of <tab> to foo is found
> under 'Global Bindings'. So, we have the old local binding (major
> mode) of TAB to lisp-indent-line that shadows a global binding of
> <tab> to foo. Now pressing the tab-key or C-i should mean the same
> thing (produce ascii code 9), which in turn should execute lisp-indent-
> line (since local bindings takes precedence over global bindings),
> right?
>
> Wrong. Pressing the tab-key executes foo, and pressing C-i executes
> lisp-indent-line. So, TAB seems to mean pressing C-i, which apparently
> differs from <tab> which seems to mean pressing the tab-key. Now, this
> is becoming a mess.
>
> 1) Shouldn't pressing the tab-key and C-i mean the same thing?
> 2) Shouldn't all emacs' representations of a tab (<tab>, TAB, ?\t,
> "\t", ascii code 9, <C-i>, ?\C-i, ?\^i etc etc) mean the same thing?
>
> Best regards,
>
> /Stefan

Ok, I finally found the solution in Emacs Manual node 57.4.8 "Named
ASCII Control Characters". I'll be damned...

/Stefan