From: Paul Davis Subject: Re: using external MIDI device to control the computer Date: Thu, 21 Mar 2002 07:01:32 -0500 Sender: alsa-devel-admin@lists.sourceforge.net Message-ID: <200203211201.HAA13010@renoir.op.net> References: <20020321064256.83273.qmail@web21302.mail.yahoo.com> In-Reply-To: Your message of "Thu, 21 Mar 2002 01:42:56 EST." <20020321064256.83273.qmail@web21302.mail.yahoo.com> Errors-To: alsa-devel-admin@lists.sourceforge.net List-Help: List-Post: List-Subscribe: , List-Unsubscribe: , List-Archive: To: fn ln Cc: alsa-devel@lists.sourceforge.net List-Id: alsa-devel@alsa-project.org > a sequencer client ? as a userspace program ? How >could it then get the (filtered) data back for the >MIDI app into /dev/midi ? I didn't think there was a >way for a userspace program to feed "incoming" data >into /dev/midi, and therefore I thought that it would >have to be a kernel module, not a userspace pgm. the sequencer API isn't structured around /dev/midi, its structured around sequencer ports. clients have their own ports which they connect to other ports. all the raw midi devices automatically become clients with ports (which correspond to the h/w they control) when the sequencer module is loaded, so you can always send/receive data to them. a client can be a user space program or a kernel space entity. any number of clients can be connected to a port. >How can "several clients connect to the same sequencer >port " so they all see the data ? as i said, /dev/midi isn't the sequencer, its the raw MIDI device, and there isn't a way to multiplex access to it except via the sequencer. --p _______________________________________________ Alsa-devel mailing list Alsa-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/alsa-devel