Terminal TAB key and character (Re: [vimpulse] Bug: TAB doesn't autocomplete in minibuffer.)

Sat Jul 3 12:23:41 CEST 2010

Vegard Øye <vegard_oye at hotmail.com> writes:

> On 2010-07-02 17:57, Štěpán Němec wrote:
>
>> On Fri, Jul 02, 2010 at 04:55:04PM +0200, Vegard Øye wrote:
>>
>>>
> http://www.gnu.org/software/emacs/manual/html_node/emacs/Named-ASCII-Chars.html
>>
>> ... with the the caveat that in the terminal (emacs -nw) there is no
>> such distinction (as the section you quote also mentions).
>
> I tested it and it still worked, so I guess Windows XP doesn't have an
> "ordinary ASCII terminal". The section also states that "most modern
> terminals" recognize the distinction.
>
> Štěpán, do you have access to an ASCII terminal to experiment with?
> Does the order of binding [tab] and C-i matter?

Well, IMHO "ASCII terminal" might be a little bit of a misnomer nowadays
(most terminals seem to handle non-ascii just fine), but it's a fact
well known and recognized that in a terminal TAB key sends a TAB character.

So when I run Emacs in a terminal (i.e. emacs -nw), binding [tab] has no
effect whatsoever, and binding C-i affects both the TAB key and the C-i
combo.

Are you saying that when you do C-h c C-i and C-h c <tab> in emacs -nw
under Windows, you get different results? How does emacs -nw on Windows
even look like? You're launching it inside cmd.com?

I'm not sure what the "most modern terminals" mentioned in the text
mean. I suspect "terminal" is just a synonym for "computer" there. If
there's a way to somehow re-program a text terminal to send something
different for <tab> and C-i, then I'm yet to hear about it (it might
be possible, I'm no expert on terminals), but you certainly can't rely
upon such behaviour.

> (Sorry about the separate posts.)

I see nothing wrong 'bout that.

Štěpán