From mboxrd@z Thu Jan  1 00:00:00 1970
Message-Id: <200406181606.i5IG6BW06986@zamenhof.cs.utwente.nl>
To: Fans of the OS Plan 9 from Bell Labs <9fans@cse.psu.edu>
Subject: Re: [9fans] 9term echo-disabling fix (for ssh, vncviewer)
In-reply-to: Your message of "Fri, 18 Jun 2004 17:04:24 +0200."
	<013a01c45545$8bf80c50$07fa7d50@SOMA> 
References: <200406181311.i5IDBLB05307@zamenhof.cs.utwente.nl> 
	<013a01c45545$8bf80c50$07fa7d50@SOMA>
From: Axel Belinfante <Axel.Belinfante@cs.utwente.nl>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 18 Jun 2004 18:06:11 +0200
Topicbox-Message-UUID: a62061fc-eacd-11e9-9e20-41e7f4b1d025

> > - return (ttmode.c_lflag&(ICANON|ECHO));
> > + return (ttmode.c_lflag&(ICANON) && ttmode.c_lflag&(ECHO));
> 
> that's weird.  1st returns either:
> 
>     ICANON
>     ECHO
>     ICANON|ECHO

> 2nd returns 0 [raw | raw and no echo] or 1 [not raw and echo].

the return value is interpreted as boolean,
as in your ttyechoing routine (thanks for the pointer)
(but the != 0 test is omitted here).

for some reason (why? likely because matty's code did. why?)
we are not only looking at ECHO but also at ICANON .
Then, we must be careful to correctly deal with the case
where we have [icanon and no echo] which is what
ssh and vncviewer use when prompting for a password
(shows up as: israw c- in the debug 9term output)

The above seems to do the right thing in that case.
The original would interpret [icanon] as a reason
to echo, which is incorrect (or at least, does not
hide the password that is typed).

However, I guess with the above 9term does the wrong
thing in the case where we have [no icanon and echo],
because also then it does not echo.
However, I'm less concerned about that.

However, I'm the first to admit I know very little about termio(s).

Axel.