Disabling Quit On Telnet

[Jay R. Ashworth]

On Sat, Apr 29, 2006 at 04:24:49AM -0400, Fairlight wrote:
> 1) The handling of the "X" button event is only really possible to do at
> the client level.  That isn't anything Scott will be able to do without the
> vendor changing the client.  However, I know for a fact that you can tell
> the Windows wm not to draw those three buttons -at all-.  If that were
> done, there's no way to click on them, obviously.  A feature you might
> consider from even an aesthetic standpoint--why have decorations there that
> are rendered unusable anyway?

The *proper* solution there, from a UI design standpoint, is to draw
the widgets, grey them out, and put a tooltip on them *explaining why
they're greyed out right now*.  That last step is critical, and almost
always overlooked, because it can make your state machine a cast-iron
bastard to maintain.

> 3) My best answer for server-side is to use screen(1).  You didn't specify
> your server environment, and I'm not sure anyone ever ported it to SCO.  If
> you're on Solaris, Linux, or FBSD, you're in luck.  The screen application
> will actually detect a disconnect and keep everything running--then let you
> reattach later.  Something could be pretty easily cobbled together to allow
> testing for the presence of a socket and auto-resume at login, assuming one
> is only logging in one window at a time.  One would actually only -need- to
> do so if one used screen, as it's a multiplexer and lets you have pretty
> much unlimited (more than even I use, and my default is 10 screens--8 as
> fairlite, 2 as root, and I can still go far higher).

This is, in fact, an excellent solution, and screen is sufficiently
configurable to make it pretty easy to train (where by that, I mean
"You Won't Have To").  I had this running with putty on a client site
where we remapped the screen-switch keys to Alt-Fn, which putty will
actually send -- though you have to remember to disable Alt-F4 :-) --
and I even had it prelaunch certain dscreens in certain windows...
<rant>
though I had to put a wrapper around them to release the session,
because unlike FacetTerm (where filePro makes special concessions),
each screen for a user was counted as a separate license count, and *64
users* wasn't enough for 9 human beings.
</rant>

> So yes, it works.  Actually, that's a filePro <-> SCO 5.6 <-> screen <->
> linux <-> PuTTY on Win2k solution I'm using in that particular scenario. :)

And it works quite nicely, assuming you get all the emulation settings
correctly matched.

> The nice thing is, it protects you from just about anything.  The power
> could go out and the UPS could shut down your computer--you could still
> reattach.  It doesn't matter -why- you were detached, you can reattach.

Indeed it's good.  It's good enough, and transparent enough, that lots
of *nix people wire it right into their .profile.  For some reason,
I've never been one of them; haven't gotten burned enough times, I
guess.

Cheers,
-- jra
-- 
Jay R. Ashworth                                                jra at baylink.com
Designer                          Baylink                             RFC 2100
Ashworth & Associates        The Things I Think                        '87 e24
St Petersburg FL USA      http://baylink.pitas.com             +1 727 647 1274

     A: Because it messes up the order in which people normally read text.
     Q: Why is top-posting such a bad thing? 
     
     A: Top-posting.
     Q: What is the most annoying thing on Usenet and in e-mail?