Last Friday 11 March 2005 09:41, ish(a)sarai.net was like: I remember there being a version in A/DeMuDi with JACK support, but I never had any success getting it to work. I'm not sure whether it's really being maintained at the moment. cheers, tim hall http://glastonburymusic.org.uk

I take back what I said about compiling Bristol. I built the buildable bits and moved the whole directory into /opt (or you could edit the startup script and put it somewhere else). Then, following the instructions in the thread mentioned below, I put "jacklaunch" before the 2 occurences of "bristolengine" at the end of the startBristol script and I now have a working Jackified Bristol. Well kind of. The audio engine and the jack routing work, but the midi doesn't. Or at least it doesn't show up in the patchbay in qjackctl. Any ideas on how to get midi working with Bristol and libjackasyn? Is it a question of doing something to Bristol or is it something to do with how I configure alsa or my kernel? Thanks Robert On March 11, 2005 10:31 pm, quoth Robert Persson: -- Robert Persson "No matter how much ye shake yer peg The last wee drap rins doon yer leg."

Tobias Ulbricht wrote: The MIDI Connections patchbay in QJackCtl is a nice convenience, it's not a fundamental aspect of JACK itself. The MIDI patchbay represents (IIRC) the status of the ALSA sequencer. If Bristol is not an ALSA sequencer client then it will not display in the MIDI patchbay, though as a JACK client it will appear in the Audio Connections tab. As far as I know, at this point JACK has nothing to do with MIDI. Best, dp

I am beginning to see that, whatever the problems with Bristol may or may not be, there are some basic things about how midi is routed in Linux that I need to understand first. i.e. there are a significant number of different files in /dev that are midi related, but I don't really know the difference between all of them. For instance I have not been clear about the difference between midi, raw midi and the sequencer (it is only from reading Dave Phillips' post in this thread that I even know that it is the sequencer we are dealing with in the midi patch bay), nor have I been clear whether, for instance, /dev/sequencer is an alsa or oss device. Is there a guide somewhere that explains this kind of stuff? Another thing: Am I right to understand that whether or not I use jacklaunch should in itself make no difference to what happens with midi? In other words will "jacklaunch ./startBristol" and just "./startBristol" do the same thing? Experiment seems to tell me that, but I want to make sure. Anyway, this is what happens when I try to launch Bristol in various ways: $jacklaunch ./startBristol -midi oss $jacklaunch ./startBristol -midi alsa Both of the above cause Bristol to launch with no midi-related error messages and no midi in the patchbay. $jacklaunch ./startBristol -seq This is supposed to start Bristol using the alsa sequencer. The application tries to start up, a window opens but no gui appears, "129:Bristol" shows up for a second or so in the midi patchbay and then Bristol shuts down. Here is the output bash-2.05b$ jacklaunch ./startBristol -seq You may want to make bristolengine a suid-root executable spawning midi thread Could not reschedule thread to 2 parent going into idle loop connected to :0.0 (80c53d8) display is 2752 by 1200 pixels Window is w 2752, h 1200, d 24, 0 0 0 Using DirectColor display masks are ff0000 ff0000 ff0000 flags are 8a000000 midi sequencer Returning socket 5 Opened listening control socket: 5028 Client ID = 129 Queue ID = 0 Device name did not parse, defaults 128.0 Cannot subscribe port 0 from client 128: Operation not permitted Error opening midi device /dev/midi, exiting midi thread INIT: 80c50f0 Initialise the mini link to bristol: 80c9e50 hostname is localhost, bristol port is 5028 Connected to the bristol control socket: 5 bristolengine already active 80c4f40 80000000 0 parent exiting return - no data in buffer cleanupBristol(0) ./startBristol: line 127: 8259 Broken pipe bristol $* -engine bash-2.05b$ Does that make things any clearer? There are other options I could mess about with, such as -mididev, but those seemed most obviously connected. Robert On March 14, 2005 04:30 am, quoth Dave Phillips: -- Robert Persson ireneshusband(a)fastmail.fm YahooMess:ireneshusband AIM:shamanicpolice "No matter how much ye shake yer peg The last wee drap rins doon yer leg."

On March 14, 2005 02:48 pm, quoth Sean Bolton: Your last post introduced me to virmidi, which was something I didn't know about before. The useful thing I have learned from this latest post is that the mididev is the midi source that bristol listens to, rather than the bristol engine itself. This means that I now have Bristol sort of working, but there are still some problems: 1. No pitch wheel control via midi. 2. Dropped notes while using midi. Enough to make Bristol unusable. 3. Hammond emulator doesn't produce sound and crashes very quickly when you start fiddling with it. The other emulators seem to be alright (although there are a couple I haven't tried yet). Do you get the same dropped notes problem with virmidi? If not, then while you're trying my solution I think I'll try yours :) Robert -- Robert Persson ireneshusband(a)fastmail.fm YahooMess:ireneshusband AIM:shamanicpolice "No matter how much ye shake yer peg The last wee drap rins doon yer leg."