Re: [Nano-devel] more on Home/End not working atstatusbarw/ALT_KEYPADon xterm

[Wouter Van Hemel]

(lots of random snipping)

On Sun, 13 Jul 2003, David Lawrence Ramsey wrote:

> I don't know how to fix this.  At least you can use Esc as a workaround
> in the meantime.
>

Yes, that works fine.

> From this, it appears that more is broken when using -K than when not
> using it, and the breakage without -K is just the Backspace/Delete
> problem.
>

Exactly.

> >> 1. With the hack applied, do all the keys work as expected?
> >
> >backspace doesn't work (deletes to the right, like Delete).
> >
> >PageUp/PageDown/Insert don't work in keypad.
>
> Are these in the console or xterm, or both?
>

This is for both patched and unpatched, sorry.

xterm, _numeric keypad_ PageUp/... keys, in editor and prompt.

(NumLock glitch detected.  Keypad will malfunction with NumLock off)

NumericHome generates 'w', NumericEnd generates '\',
NumericPgUp generates 'y', NumericPgDn generates ']',
NumericUp generates 'x', NumericDown generates '^',
NumericLeft generates 't', NumericRight generates 'v',
and  '5' generates 'u', for sake of completeness. :)

This is the xterm-numpad-glitch problem, not related to your patch, I
guess.


The only difference I see with your (v3) patch is that Home and End work
in prompt (bottomwin) on freebsd console, with '-K'. It doesn't break
anything on freebsd. So if it doesn't break other platforms, it's ok.

But I don't understand why I would use -K since it doesn't fix anything
(from a freebsd point of view), and even breaks things without your patch.

> As if it wasn't enough of a mess already.
>

Yeah. :(

> >The only problem is that some keys (PgUp, PgDn, Ins) don't work good in
> >a nano prompt, alt fails in console, and then Delete/BackSpace. I
> >suggest a simple argument that just switches these two keys, like pico
> >had IIRC.
>
> That would be the -d option.
>

In short:

Only considering freebsd, I'd say ditch -K and have a simple -d to switch
the %€#"@! Delete/BackSpace key.

Your patch fixes some of the -K issues (so it's good).

-K stops the 'S', 'P', 'K' from shown in console term, nano prompt
(because it uses other bindings, I guess), but shows ~ instead of 'S',
'P', 'K' in xterms, nano prompt. On FreeBSD, it's thus quite useless,
except for the switching of Delete/BackSpace:

  nano   , console: Delete/BackSpace work fine
  nano -K, console: Delete becomes BackSpace
  nano   , xterm  : BackSpace becomes Delete
  nano -K, xterm  : Delete/BackSpace work fine

Both in the editor window and prompt window, with or without your patch.

> >The others are quite trivial, I think those specific codes are just not
> >implemented in the nano prompt.
>
> You're right.  The keys don't do anything at the prompt when pressed,
> but nano doesn't block such unused keys (cf. the comment in nano.c about
> how "we no longer stop unhandled sequences so people with odd character
> sets can type").  Also, the extended keycodes generated when you press
> those keys (when using -K) are outside the char range, so when they get
> cut down to char range just before being displayed, you get 'S', 'P',
> and 'K'.  It's odd but harmless.
>

Without -K the last part of the sequence is shown ('S', 'P', 'K'), in
console and xterm. With -K console shows nothing and xterm shows '~'.

It's not important, as you say, because those keys aren't needed anyway.


Thanks for your efforts,

  wouter