Re: gold key for edt mode

["G. Milde" <g.milde@xxxxxx>]

Hej Thei,

On 30.01.06, Thei Wijnen wrote:
> I have some comments on your reply:

> 1. x-keydefs for / * - and +
>   If I put the following in jed.rc these keys work as in jed-EDT.
> #ifdef XWINDOWS
>   require("x-keydefs");
>   x_set_keysym(0xFFAA, 0, Key_KP_Multiply);
...

This is basically what x-keydefs does (in the code below the #stop and in a
manner that it can be used with xjed and jed also if preprocessed).

> I changed xkeys.c to achieve the same and still wonder whether this
> can be done in the next code release.

* In previous tests, this seemed to be consistent with the output in an
  x-terminal-emulator (using Ctrl-V in rxvt).

* Now, (maybe after the upgrade to X.org) the numeric keypad in rxvt
  consistently sends "what is printed on" if NumLock == On and
  \eO<something> sequences if NumLock=Off.

* However, 
    keysyms for KP_Delimiter and KP_0 ... KP_9 change with NumLock state 
  while 
    keysyms for KP_Divide ... KP_Enter are independent of NumLock state
    (i.e. xjed has no chance to distinguish them by the keysym).

* By default (without x-keydefs) we have in xjed

    Key_KP_Return    -> "\eOM"
    Key_KP_Divide    -> "/"
    Key_KP_Multiply  -> "*"
    Key_KP_Add       -> "+"
    Key_KP_Subtract  -> "\eOm"
    Key_KP_Separator -> "\eOn" % [, Del]  if NumLock==0, 
    Key_KP_Delete    -> "\eOl" % [, Del]  if NumLock==1
    
    Key_KP_0         -> "\Op"
    ...
    Key_KP_9	     -> "\Oy"
    
  regardless of the NumLock state (besides the noted exception).
  
  
This is pretty inconsistent and I support a call for change in
xkeys.c

FEATURE REQUEST: consistent default keybindings for the keypad in xjed

  Variants:
    
    1. Return VT220 sequences ("\eO<something>") for all numpad keys
       regardless of the NumLock state.
       
    2. Return VT220 sequences ("\eO<something>") for all numpad keys
       if NumLock is Off
       
       Return "0" ... "92 for KP_0 .... KP_9 and and "," for KP_Delete
       if NumLock is On (should be easy, as these keys have different
       keysyms).

    3. If the NumLock state can be queried from within xjed:
    
       Return VT220 sequences ("\eO<something>") for all numpad keys
       if NumLock is Off
       
       Return the symbol that is printed on for all numpad keys
       if NumLock is On
       
       Provide a symbol for querying the NumLock state in x_set_keysyms,
       say
          x_set_keysym, e.g.        
          0   : unmodified key
          '$' : shifted
          '^' : control
          'n' : NumLock
       and add documentation for 
          '%' : shift-control   

> 2. The standard edt.sl contains some of the smartscroll code
> (on cursor keys only) and that leads to unexected jumps of
> the screen/cursor when other keys bring you to the screen border.
> That is why a added my extra code.
> By the way there is a simple way of turning this feature off
> by making the variable  EDT_Scroll_Begin public.
> Then in jed.rc you can set it to 0 to get "no smartscroll".
> 
> Providing a smartscroll.sl is another option we could work on
> but then I would like to remove the partial solution from edt.sl.

John, could we remove the partial "smartscroll" code from edt.sl (or make
it configurable with e.g. EDT_Scroll_Begin?)



Thei, answering would be a lot easier for me if you did not use "TOFU"
(Text Oben Fullquote Unten = text above full quote below).


Günter


--------------------------
To unsubscribe send email to <jed-users-request@xxxxxxxxxxx> with
the word "unsubscribe" in the message body.
Need help? Email <jed-users-owner@xxxxxxxxxxx>.