1.  Do we *know* that external 7.1 does not currently work, with dotdeb#29 installed?  I have not tested it.  Is there a vu-meter app that will show me when external 7.1 jacks are being correctly exercised, without needing to plug in all the physical 3.5mm wires that actual 7.1 external testing would typically necessitate?  

2.  dotdeb#29 and dotconf#13 are both installed, and seem to be working -- I am a little worried that my next ubuntu update will wipe them out, however (the dotconf change seems especially likely to suffer this fate).  

3.  I'm not 100% positive the subwoofer is failing to woof... all the speakers are within a few inches of each other.  But yes, I currently believe that 'LFE' which is now verbalized as 'rear center' is incorrectly outputting through the physical '(front) Center' speaker, instead of the desired physical subwoof.   Do you have some sample 5.1 audio-files which exercise each speaker in turn?  Or is that no different than speaker-test with six channels specified....   

4.  Should I 'unapply' the comment#32 workaround, or no need to mess with that? 

5.  I need to delete ~/.pulse/pulse.conf so that pulseaudio will be automatically active when the machine boots.  

6.  I am willing to compile an alsa driver, or whatever, if that will help Raymond test his theories.  I'm not sure how to make a custom dkms dotdeb package of my own, however.  See also point one above -- maybe it is already working, we just don't know it yet. 

7.  Raymond mentioned the bug where plugging in the hdfons fails to auto-mute the internal speakers; that was marked as fixed in 12.04, and indeed it works for me with dotdeb#29 installed (and also worked before).  

8.  Raymond thinks that 'PCM' for out-node 0x05 is a dumb name... and I tend to agree.  Back before workaround#32, however, when the machine first booted it was defaulting to using that node as the default audio-sink, and then sound-settings was getting me stuck with node 0x06 aka spdif which is not plugged into anything as the 'selected' new default audio-sink.  Following the dark lines in codecgraph's SVG output, node 0x05 pcm_out is connected via mixer#0x0f over to pin#0x17 which is labelled "jack line out at ext rear stereo, in out detect trigger".  Is that a problem?  

0x02_out = speaker front , mix 0x0c, pin 0x14 fixed speaker at int atapi hp eapd , recursive via mix 0x0b/0x22/0x23
0x03_out = speaker clfe , mix 0x0d , ping 0x15 fixed spkr at int atapi , recursive via mix 0x0b/0x22/0x23
0x04_out = speaker surround, mix 0x0e , pin 0x16 fixed spkr at int atapi , recursive via mix 0x0b/0x22/0x23
0x05_out = pcm , mix 0x0f , pin 0x17 jack line out at ext rear , recursive via mix 0x0b/0x22/0x23
pin 0x1A line in at ext rear , connects 'recursively' to the main four outputs via mix 0x0B 
pin 0x18 mic at ext rear , connects 'recursively' to the main four outputs via mix 0x0B 

0x06_digital_out = spdif iec958 , no mix shown , pin 0x1e jack spdif out at ext rear digital 
0x10_digital_out = noName_notEnabled , no mix shown , pin 0x11 N/A spkr at ext rear digital 
0x25_out = headphones , mix 0x26 , pin 0x1B jack hp out at ext rear hp eapd 
0x09_in = capture , mix 0x22, pin 0x12 fixed mic at int atapi 
0x0A_digital_in = noName_notEnabled , no mix shown , pin 0x1F N/A spkr at ext rear digital 
0x08_in = noNameMaybeEnabled? , greyed connection to mix 0x23 , greyed connection to pin 0x19/0x1c/0x1d (spkr*3)

9.  Insert your issue here.