From: cgf AT kramden DOT cygnus DOT com (Christopher G. Faylor)
Subject: Re: Tty support defaults
16 Apr 1998 22:26:57 GMT
Message-ID: <6h60jh$nab$1@cronkite.cygnus.com>
References: <Pine DOT SUN DOT 3 DOT 91 DOT 980415081846 DOT 22419A-100000 AT tweedledumb DOT cygnus DOT com> <353a1939 DOT 43750771 DOT cygnus DOT cygwin32 DOT developers AT smtp DOT goodnet DOT com>
X-Newsreader: trn 4.0-test63 (15 March 1998)

I'll look into this ASAP.

In article <353a1939 DOT 43750771 DOT cygnus DOT cygwin32 DOT developers AT smtp DOT goodnet DOT com>,
Mikey <jeffdbREMOVETHIS AT goodnet DOT com> wrote:
>Hi Chris ;^)
>to reproduce the bug.
>
>fire up 95
>set CYGWIN32=notty (or just leave unset)
>bash
>^C
>for file in *;do rm ../$file;done #to get file not found errors
>LOCKUP!!!
>
>On Wed, 15 Apr 1998 08:21:40 -0400 (EDT), you wrote:
>
>>
>>On Wed, 15 Apr 1998, Geoffrey Noer wrote:
>>
>>> > The notty default causes needless lockups, and, I don't see
>>> > any reason why we would want to have it turned off.
>>> > comments?
>>> 
>>> It sounds like we may be moving in the direction of putting some more
>>> effort into the tty layer soon.  If we decide to this and it proves
>>> feasible, we would change things so that the tty layer was always on
>>> while still permitting non-Cygwin32 console applications to run.
>>> Currently, if tty mode is on, some console apps like the non-Cygwin32
>>> version of vim can't run from bash which is very annoying.
>>> 
>>> I completely agree that having a tty mode is not desirable in
>>> principle.  For now though it's better than having it always on, IMHO.
>>
>>I agree.  I don't understand the reason for the lockups in the console
>>mode version.  I haven't experienced them.  But, they should be fixed.
>>
>>Maybe eventually, we'll want to have some method for marking specific
>>binaries as "notty" by default.
-- 
cgf AT cygnus DOT com             "Everything has a boolean value, if you stand
http://www.cygnus.com/      far enough away from it."  -- Galena Alyson Canada