I've found that the device numbering situation is a bit more complicated 
that I had thought.

>From my error logs, I note that:
1) the  hw:x,x is expected to be that of the MIDI device 
installed. (That we already knew.)
*and* 2) the audio hardware card device # is expected to be hw:0,0 in all 
cases. (Otherwise it [its file] will not be found.)

So, if the USB MIDI device is inserted before boot, it will automatically be 
assigned hw:0,0 and regardless of other factors, Csound will fail.

I don't think that even an -Ma patch by Andres will solve this bottleneck, 
will it?

Art Hunkins

----- Original Message ----- 
From: "Peter Robinson" 
To: "Art Hunkins" 
Cc: 
Sent: Friday, September 11, 2009 2:56 PM
Subject: Re: [Csnd] Re: Re: Re: Re: Re: Re: Re: Re: Re: csound on Fedora 11 
and rawhide


> On Fri, Sep 11, 2009 at 7:26 PM, Art Hunkins  wrote:
>> I've finally gotten to the bottom of the (single) MIDI controller problem 
>> in
>> SoaS (Sugar-on-a-Stick).
>>
>> What was causing me much confusion was the fact that when the controller 
>> was
>> plugged in prior to boot, MIDI didn't work. When plugged in *after* stick
>> boot, it *did* work. (Strange, but true.)
>>
>> The trick was to consult ls -lR /dev/snd/* as several of you suggested.
>>
>> Using the : -rtmidi=alsa -M hw:1,0 in all cases:
>>
>> Inserting stick before boot gave these values in ls -lR /dev/snd/*:
>> midiC0D0 and hwC1D0
>>
>> Inserting *after* boot:
>> hwC0D0 and midiC1D0
>>
>> Switching the USB MIDI device in and out after boot, doesn't change any 
>> of
>> these data.
>>
>> Obviously, the above discrepancy is a problem.
>>
>> The log file (of failures) shows this:
>> ALSA lib rawmidi_hw.c:233:(snd_rawmidi_hw_open) open /dev/snd/midiC1D0
>> failed: No such file or directory
>> ALSA: error opening MIDI input device
>> *** error opening MIDI in device: -1 (Unknown MIDI error)
>>
>> What's the solution? It's important that both scenarios work.
>
> Well that would be, at a random guess, because there's two sound
> devices! And the devices get scanned in different orders. USB first
> which means it gets C0 on boot and the onboard would get C1 (C ==
> controller or something) but would reverse after. the fix would be to
> have a wrapper script that checks for devices (multiple midi would
> prob be C2 etc - that's your hint for your other fix) and then set a
> variable and specific it in the command line to start the app..
>
> Peter 



Send bugs reports to this list.
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"