Mailing-List:	contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe:	<mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive:	<http://sourceware.org/ml/cygwin/>
List-Post:	<mailto:cygwin AT cygwin DOT com>
List-Help:	<mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs>
Sender:	cygwin-owner AT cygwin DOT com
Mail-Followup-To:	cygwin AT cygwin DOT com
Delivered-To:	mailing list cygwin AT cygwin DOT com
To:	cygwin AT cygwin DOT com
From:	Andrew DeFaria <Andrew AT DeFaria DOT com>
Subject:	Re: Running cmd programs in cygwin / rxvt
Date:	Mon, 20 Sep 2004 08:31:41 -0700
Lines:	42
Message-ID:	<cimt5h$m4f$1@sea.gmane.org>
References:	<414E8B1A DOT 8040103 AT overbored DOT net> <414EA28E DOT 122EB347 AT dessent DOT net> <cimon4$783$1 AT sea DOT gmane DOT org> <414EEDAB DOT B2852F85 AT dessent DOT net>
Mime-Version:	1.0
X-Complaints-To:	usenet AT sea DOT gmane DOT org
X-Gmane-NNTP-Posting-Host:	h-67-102-25-114.lsanca54.covad.net
User-Agent:	Mozilla Thunderbird 0.8 (Windows/20040913)
In-Reply-To:	<414EEDAB.B2852F85@dessent.net>
X-IsSubscribed:	yes

Brian Dessent wrote:

> Andrew DeFaria wrote:
>
>>> What you're seeing is caused by the output of the program going 
>>> through a tty (or pty as the case may be.) When you run it from a 
>>> straight cmd.exe it's not using tty code (unless you have CYGWIN=tty 
>>> set.) From my meagre knowledge of the subject it has to do with the 
>>> program not explicitly flushing stdout (or not knowing that it needs 
>>> to as this is usually not required under windows.)
>>
>> Is there a CYGWIN setting to tell rxvt not to use ptys?
>
> I don't think that would do any good. rxvt is a posix program and 
> expects an environment where it has ptys that it can use. Without pty 
> support it would likely complain loudly and refuse to run.
>
> Again, I'm pontificating on something I know little about here, but 
> I'm guessing one solution might be some kind of kludge in the cygwin 
> layer that would attempt to detect when a native app is doing buffered 
> writes to a pty and autoflush after each call. But, just a WAG. It 
> might even be out of cygwin's control, if the buffering is done in the 
> CRT of the app.

Cygwin does do some of it's own coding that is conditionally compiled 
right? Why not conditionally compile stuff to get ptys (or some other 
equivalent) to function with native Windows apps that are currently 
having problems with the concept of ptys?

I know I often point people to use rxvt because it's a lot better than 
that silly Windows window. However I often work with Clearcase and 
Clearcase's main command, cleartool, messes up WRT rxvt and ptys (as 
well as other tools), so I'm constantly explaining the problem. I'd just 
like a real solution to offer and to use.

I don't understand the internals of ptys vs. ttys and native console I/O 
nor why it's having the problem. I just deal with the result and would 
like it fixed if possible. Sounds like it may be a complex fix though. 
As I said I'm not sure why ptys are necessarily needed in the first 
place but admittedly I have not researched it.
-- 
Few women admit their age. Few men act theirs.


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/

- Raw text -