Mailing-List: contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-developers-owner AT sources DOT redhat DOT com Delivered-To: mailing list cygwin-developers AT sources DOT redhat DOT com Message-ID: <001b01c0b397$2d1ca720$0200a8c0@lifelesswks> From: "Robert Collins" To: Subject: setup will have to wait :[ Date: Fri, 23 Mar 2001 23:45:25 +1100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 X-OriginalArrivalTime: 23 Mar 2001 12:40:08.0421 (UTC) FILETIME=[64D04150:01C0B396] I just ran into another apparently missing function - mkfifo. I've thought up a way to implement this - but I'd like some feedback from someone with more machines that openBSD & win2000. what's the expected result of mkfifo foo. cat foo > 1 & cat foo > 2 & echo "test" > foo should I code fifo's so that the result of that test will be cat foo > 2 returns a read error (file locked), or so that multiple reads do occur? It's much the same to me either way - I just don't know the expected behaviour of a fifo in that circumstance. For those interested in the "how" side, I'm planning on avoiding named pipes (due to the horror stories I've heard about win 9x). Instead I plan on using a manual reset named event object and a named mutex. The mutex will control access, and the event object will control syncronisation. (I.e if process 2 opens & tries to write, it will successfuly get the mutex, and then wait for the event object to be pulsed. It will then "write" the data. The recieving process will recieve the mutex when the first process completes the write. That then gets repeated. For data transfer, I was thinking of using shared named memory:anyone know of caveats for that on win9x? Rob.