Re: XPL shortcuts etc. [long]

[Robert Holmgren holmgren@xxxxxxxx]

** Reply to note from nsivin@xxxxxxxx (Nathan Sivin) Wed, 8 Jan 1997 08:58:34 -0500 (EST)
> 
> Re NI, it would be good to have a little more theoretical analysis if
> someone is in the mood. I have been struck (using DOS 6.2 for XY and
> WFW 3.11 for XW) how few of the NI's in the standard keyboard are
> necessary. In general they don't seem to be necessary when not
> followed by a command rather than ASCII characters. On the other hand,
> I have found in customizing the keyboard that they are sometimes
> necessary, often in an odd place in the sequence, to make something work.

I don't know the theory -- but the practice is, that if a
XyWrite keystroke wakes up the underlying OS, then you need to make a
choice. If you don't want that behavior, you use NI, which attempts to
erect a wall between XyWrite and the OS, preventing keystrokes from
bleeding through. Otherwise, your best bet is simply to comment out the key
entirely (with a semi-colon), and let the OS have the key. So it
depends, as somebody said, entirely on what the OS responds to. Ctrl-Esc
is a good example. It switches between running tasks in OS/2 and Win95, so
you want the OS to have it. In fact, it may have such a high
preemptive priority that NI isn't able to insulate it (I forget). Anyway,
comment it out in .KBD:
TABLE=CTRL
;1=
..
That way XyWrite doesn't intercept or even acknowledge the key; it goes
straight to the OS. Basically, you don't need NI except on keypad keys
71-83, and on some extended keyboard keys >84. And even then, the OS
is looking mostly for shifting keys, like Alt and Ctrl -- never lower case
alphanumerics.

-----------
Robert Holmgren
holmgren@xxxxxxxx
-----------