On Mon, 22 Jan 2001, Sean Porterfield wrote:
> Scott Klement wrote:
> >
> > On Mon, 22 Jan 2001, Sean Porterfield wrote:
> >
> > > When FER is set, only certain keys are valid. The ones I know of are
> > > arrow keys, TAB, BACKTAB, BACKSPACE, ENTER, FIELD EXIT, FIELD +. (This
> > > was tested on a 3487.) The odd thing about the "real" terminal is left
> > > arrow which doesn't move the cursor, all other arrow keys do. I guess
> > > it moves left back into the field.
> >
> > On most terminals, once you get to the end and you've typed the last
> > digit, the cursor flashes a little differently, and you're in a mode where
> > you need to press a field-exit (or similar key) before you can continue.
>
> Yes, I suppose that is correct. The only FER field I can think of in
> our software is a one character field, so it's hard to tell some things.
>
> > Can you elaborate on what's wrong with FIELD+? If it happened in the last
> > few weeks, its probably my fault. :( What does it do wrong?
>
> Input Inhibit goes on instead of FER off (FER stays on according to the
> display, don't know about code.)
My first thought upon reading this was "hmmm.. that doesnt happen here!",
but then I decided to review my FreeBSD keyboard mapping vs. the Linux
one. Here is what I found:
1) In the freebsd keyboard map (or, the updated version that only exists
on my system at present) fieldexit is mapped to ctrl-k, and
fieldplus is mapped to ctrl-x.
2) The linux version doesn't map a field plus key. It's still using the
original tn5250 strangeness where the "+" key in a signed numeric
field acts as field plus, but in a character field it acts as the
"+" character.
3) In cursesterm.c, NOTHING is wired to the "real" fieldplus key. (But
both C-k and C-x are mapped to FIELDEXIT)
Therefore, I submit that the reason you're getting II in a FER field is
because you're actually sending "+", which isn't an exit key for FER.
The fix is to change your keyboard map to send ctrl-x when you press the
keypad's "+" key (which probably shouldn't go into CVS for a console mode
keyboard map... but maybe should for an xterm), and to change cursesterm
to wire ctrl-x to K_FIELDPLUS (which must be a bug).
>
> > > I added K_ENTER to my display.c and it seems to work well. I don't know
> > > how to fix up arrow, right arrow, down arrow, FIELD+, FIELD- (if FIELD-
> > > should really work - it doesn't on my "real" terminal).
> > >
> >
> > Just above where you added the K_ENTER there is some code that says
> > something like:
> >
> > case K_LEFT:
> > case K_BACKSPACE:
> > tn5250_display_indicator_clear (This, TN5250_DISPLAY_IND_FER);
> > return;
> >
> > Try adding some code below that which looks like this:
> >
> > case K_UP:
> > case K_DOWN:
> > case K_RIGHT:
> > tn5250_display_indicator_clear (This, TN5250_DISPLAY_IND_FER);
> > break;
> >
> > (note the "break" instead of "return" will cause it to move up/down/right
> > as well as clearing FER)
> >
> > Is that what you're looking for? (I'd get up and try it on a terminal,
> > but I've just spent too much time with tn5250 today, already)
>
> Keeping in mind that I'm not a C programmer, and the only one I really
> cared about was enter (there were others I cared about, but they've been
> fixed for a while)...
>
> Having said that, the above code seems to make tn5250 function as I
> would expect on my FER field. (I only have one 1 character FER field in
> the software we use...)
Weren't you the one who asked how to fix up, right, down, etc?
>
> Scott, wanna get Jay a patch? It's mostly your code... :) Otherwise
> I'll give it a shot.
I'm confused... did we decide to change the behavior of the arrow keys
or not?
+---
| This is the LINUX5250 Mailing List!
| To submit a new message, send your mail to [EMAIL PROTECTED]
| To subscribe to this list send email to [EMAIL PROTECTED]
| To unsubscribe from this list send email to [EMAIL PROTECTED]
| Questions should be directed to the list owner/operator: [EMAIL PROTECTED]
+---
