Mailing-List: contact cygwin-developers-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-developers-owner AT cygwin DOT com Delivered-To: mailing list cygwin-developers AT cygwin DOT com Date: Sat, 24 Aug 2002 15:10:35 -0400 From: Christopher Faylor To: cygwin-developers AT cygwin DOT com Subject: Re: readv/writev Message-ID: <20020824191035.GB27153@redhat.com> Reply-To: cygwin-developers AT cygwin DOT com Mail-Followup-To: cygwin-developers AT cygwin DOT com References: <00c501c2496e$39ae2720$6132bc3e AT BABEL> <1029976536 DOT 27825 DOT 46 DOT camel AT lifelesswks> <025401c24a0d$1c996380$6132bc3e AT BABEL> <1030060087 DOT 8462 DOT 14 DOT camel AT lifelesswks> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1030060087.8462.14.camel@lifelesswks> User-Agent: Mutt/1.3.23.1i On Fri, Aug 23, 2002 at 09:48:06AM +1000, Robert Collins wrote: >Yah. One thing I've been pondering for a while - strip out all the >text/binary translation for non FILE calls. I have done *any* research >to see if there are text-dependent apps using fd based IO, but there >probably are :p. Well, yeah. Pipes certainly can be either. It may only be pipes, though. I've recently removed the ability to do this with console and (I think) ttys. It shouldn't be necessary for serial I/O either since there are already existing UNIX mechanisms for dealing with this. There shouldn't really be that much overhead to the current method for handling text/binary reads, though. It's just checking a bit field. The handling of setting whether a fhandler is binary or text is still hairy, though. Every time I try to change it, I'm amazed by how complex the rules are. cgf