Re: Destructive backspace, keyboard mapping and displaycolors

I wrote some patches to address this very issue when I first started using
the program.  Another big issue, for me at least, was constant lock ups of
the client whenever I would enter invalid text in certain fields while
mid-session.  Around line 42 of display.c, I commented out this suspect
branch of code, which solved the problem:

   /* FIXME: Translate from terminal key via keyboard map to 5250 key. */

   if (tn5250_display_inhibited(This)) {
      if (key != K_SYSREQ && key != K_RESET) {
       tn5250_display_beep (This);
       return;
      }
   }

-brian


On Tue, 22 Oct 2002, Erik Zetterberg wrote:

> Hi,
>
> In brief. Yes, I agree with most of your comments. I was considering a
> config option or at least a compilation define for the destructive backspace
> but saw the task going through the ceiling timewise. Please understand that
> the changes were made to adress specific requirements
> in the shortest time possible given the knowledge we had. For us the
> constraints were to achieve a 'poor mans' Client Access if you will.
> As far as I know none of our users has ever seen a real IBM 3179 or 525x
> terminal so that was not a problem for us.
> More comments inserted after yours below.
>
> >
> >Hi,
> >
> >Here are some comments/ideas about your patch:
> >
> >[SNIPPED reason for changes]
> >
> > >     case K_BACKSPACE:
> > > +   /* destructive backspace */
> > >        tn5250_display_kf_backspace (This);
> > > +      tn5250_display_kf_delete (This);
> > >        break;
> >
> >This will work, but what about those of us who don't want destructive
> >backspace?  This should be a config option so that each user can turn
> >it off & on as required.
> >
> >[SNIP]
> >
> > > +   /* If in first position of field, stay there */
> > >     if (tn5250_display_cursor_x (This) == tn5250_field_start_col (field)
> >&&
> > > tn5250_display_cursor_y (This) == tn5250_field_start_row (field)) {
> > > -      field = tn5250_display_prev_field (This);
> > >        if (field == NULL)
> > >    return; /* Should never happen */
> > >        tn5250_display_set_cursor_field (This, field);
> > > -      if (tn5250_field_length (field) - 1 > 0)
> > > -  tn5250_dbuffer_right (This->display_buffers,
> > > -        tn5250_field_length (field) - 1);
> > >        return;
> > >     }
> >
> >Once you've removed the "field = display_prev_field (This)" the
> >"if (field==NULL)" is superfluous, and should also be removed.
>
> Anything that says 'should never happen' stays :-) One day it will according
> to Murphy. But seriously though; Yes.
>
> >
> >But, again, this should be a configuration option, so that users can
> >turn it on or off depending on their preference.   For example, I want
> >my cursor to go to the previous field when I hit backspace.  I'm used
> >to RUMBA, MochaSoft and standard IBM greenscreen terminals, all of which
> >backspace to the previous field.
> >
> >So, if you want backspace to stop, I want it to go to the previous field,
> >how do we accomodate both of us?   With a config keyword.
> >
> >[SNIP]
> >
> > > +        <Key>BackSpace: string("\177") \n\
> >
> >This breaks the BackSpace key on my computer.   After applying this patch,
> >BackSpace works exactly like Delete does.   In fact, my ASCII chart says
> >that \177 is "delete".   \010 is Backspace.   But, maybe that's a
> >difference between Sweden and the US?
> >
> >At any rate, without this change, destructive backspace works as I
> >expected it to.   Can try using this instead:
> >
> >             <Key>BackSpace: string("\010") \n\
> >
> >If that works for you, we could put that in the main distro, and it
> >SHOULD work for everyone.
> >
> > > + ~Shift <Key>Insert: string("\033[2~") \n\
> > > + ~Shift <Key>Delete: string("\033[3~") \n\
> >
> >These are the escape codes that Insert and Delete send on my system (by
> >default.)  So this works fine on my system (but it worked without doing
> >this as well)
> >
>
> >[SNIP]
> >
> > > +! Color scheme which is easier on the eyes than the default
> > > +xt5250*color0: seashell
> > > +xt5250*color1: red4
> > > +xt5250*color2: darkblue
> > > +xt5250*color3: yellow3
> > > +xt5250*color4: blue3
> > > +xt5250*color5: magenta3
> > > +xt5250*color6: blue
> > > +xt5250*color7: #888888
> > > +xt5250*color8: white
> > > +xt5250*color9: red
> > > +xt5250*color10: green
> > > +xt5250*color11: yellow
> > > +xt5250*color12: blue
> > > +xt5250*color13: magenta
> > > +xt5250*color14: darkgreen
> > > +xt5250*color15: #444444
> > > +xt5250*cursorColor:              red
> >
> >Color schemes are another one of those things that are user-preferences.
> >It's my opinion that the main distribution of tn5250 should stick to the
> >default color scheme, so that it looks like a "normal" 5250 terminal.
>
> Nolo contendere.
> The main distribution SHOULD stick to the default color scheme.
> I just lacked the examples of how to change things in the proper places it
> seems.
>
> >
> >If you want to change it, I'd recommend changing it with the color
> >keywords in your tn5250rc file instead of in XTerm, so that when you
> >update to a new version of tn5250 you don't lose your changes.
> >
>
> OK. So this is something I missed in the docs. Please direct me. I probably
> didn't realise what I was reading. When I found out how to put the relevant
> host and code-map stuff in the tn5250rc file I didn't understand that there
> was more.
>
> > >
> > > The color/colour issue really requres a user configurable option to be
> > > really user friendly. This will do for now though.
> > >
> >
> >By using tn5250rc, the individual user has the ability to override them as
> >well.
> >
> >for example, you can have systemwide defaults in
> >        ${PREFIX}/etc/tn5250rc
> >
> >and each user can have his/her customizations in
> >        ${HOME}/.tn5250rc
> >
> > >
> > > The keyboard issue is actually a mismatch between the XTerm resource
> >file
> > > and the terminfo entry supplied for 'xterm-5250'. There are more
> >mismatches
> > > in the terminfo entry ( all the F-keys) but seeing they work I decided
> >the
> > > interpretation of keypresses was being done outside terminfo/ncurses and
> > > didn't delve deeper.
> > >
> >
> >I'm thinking that this is probably a difference between the Swedish
> >keyboard and the American keyboard.  It's not a "mismatch" so much as it
> >is an assumption, in the xterm-5250 definition, that all copies of xterm
> >in all countries send the same escape codes.
> >
> >Of course, the proper fix is a full-blown keyboard mapping software, but
> >that's a big project.
>
> Yes. Been there, done that. Big! All those languages! Leave it alone!
> But again. What is in the XTerm resource file supplied doesn't match the
> terminfo supplied. I still want them to.
> BackSpace = kbs
> Delete = kdch1
> Insert = kich1
>
> >
> > > The font issue was solved by putting '-fn 12x24' in the xterm command
> >line
> > > in the xt5250 script. Can also be solved by setting the default font in
> >the
> > > XTerm resource file.
> >
> >Again, a better solution is to modify the tn5250rc file.  Again, this
> >allows you to have a systemwide default as well as a per-user default.
> >Plus, next time you install a new copy of tn5250, you won't wipe out
> >your customizations.
> >
> >To do that, add the following keywords to your ${PREFIX}/etc/tn5250rc
> >file and/or the ${HOME}/.tn5250rc file:
> >
> >                    font_80=12x24
> >                    font_132=12x24
>
> Saw that somewhere and then lost it.
>
> >
> > > If someone can direct me to a sans serif font of the
> > > same size I would be greatful though. The serifs clutter the screen.
> >
> >My only suggestion would be to run xlsfonts(1) or xfontsel(1) to
> >see if you can find a font that's more to your liking.
>
> Did that. 10x20 is the lagest monospace sans serif I can find. It is the
> 'Huge' font used in Xterm. This is a Red Hat 6.2 base. I didn't set it up
> originally though. True Type would probably be the best bet but MS has
> retracted their previously available downloadable font library.
>
> >
> >Good Luck, and thanks for the feedback!
> >
> >_______________________________________________
> >This is the Linux 5250 Development Project (LINUX5250) mailing list
> >To post a message email: LINUX5250@midrange.com
> >To subscribe, unsubscribe, or change list options,
> >visit: http://lists.midrange.com/cgi-bin/listinfo/linux5250
> >or email: LINUX5250-request@midrange.com
> >Before posting, please take a moment to review the archives
> >at http://archive.midrange.com/linux5250.
>
>
> _________________________________________________________________
> Planera semestern på MSN http://www.msn.se/resor/default.asp
>
> _______________________________________________
> This is the Linux 5250 Development Project (LINUX5250) mailing list
> To post a message email: LINUX5250@midrange.com
> To subscribe, unsubscribe, or change list options,
> visit: http://lists.midrange.com/cgi-bin/listinfo/linux5250
> or email: LINUX5250-request@midrange.com
> Before posting, please take a moment to review the archives
> at http://archive.midrange.com/linux5250.
>
>