On Saturday 04 September 2004 09:21 pm, Silvan wrote:
> On Saturday 04 September 2004 09:47 am, Chris Cannam wrote:

> Then I changed it to
>
> track 3 -> device A/connection A -> instrument/channel 3
> track 4 -> device B/connection A -> instrument/channel 3

This is what did it.  It happens that my kernel froze while burning a CD (damn 
patched kernel!  such a heavy price to pay for audio :( ) and my fixed 
version of this file never got saved.  I had saved it in the intermediate 
state, so I've had a chance to take a second look.

Where it sits right now, the scenario described above is exactly the case, and 
that's what's screwing up the programs being used.  device/connection A is my 
first emu10k1 port.  If I assign device B (normally 64:0) to 65:1, then 
everything works fine.

It happens that if I leave the above scenario in place, with A-A B-A, then I 
go in and assign the device B tracks to device A, the conflict persists even 
though I no longer have any tracks assigned to device B at all.

Anyway, unless there's a really compelling reason otherwise, I think the best 
fix to avoid this kind of trouble is to force connections to be unique.  If I 
have device A hooked to connection A and then I set connect device B to 
connection A, the result should be that device B gets connection A, and 
device A is unassigned.

I can see the value of being able to just assign multiple devices to use the 
same connection.  That's what I was trying to do all along.  But it doesn't 
work because devices are tied to instruments irrespective of connections.  
There's no way to spot and deal with a conflict, and it's entirely too easy 
to create a confusing set of circumstances.  We could deal with this in other 
ways, but it seems fairly clear to me that the path of least resistance is to 
prevent this situation from developing at all.

I'd file a bug on this, but SF is down.

-- 
Michael McIntyre  ----   Silvan <[EMAIL PROTECTED]>
Linux fanatic, and certified Geek;  registered Linux user #243621
http://www.geocities.com/Paris/Rue/5407/
http://rosegarden.sourceforge.net/tutorial/


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
Rosegarden-devel mailing list
[EMAIL PROTECTED] - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-devel

Reply via email to