[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Persistent names for audio devices.



From:	Reco <recoverym4n@enotuniq.net>, Sun, 20 Oct 2019 20:57:52 +0300
> So, does it work?

Yes.  At present dmix:CARD=Live,DEV=0 produces sound.

dmix:CARD=Live,DEV=1
and 
dmix:CARD=Live,DEV=2
do not produce sound at present.

> Try hw:CARD=Set,DEV=1. 

Will be trying it this week.  Thanks.

> The reason why dmix fails you escapes me
> currently. Certain software (like pulseaudio) can hold the sound card
> exclusively, could be the case.

Accepted.

> Yet they are predictable the same way udev makes NICs predictable.

In the recent months I observe that some days "AUDIODEV=hw:0,0 play MY/m85.WAV" 
alone produces sound and other days "AUDIODEV=hw:1,0 play MY/m85.WAV" 
alone produces sound. We should say the name is predictable but function is not.  

I know about udev and NICs.  For years udev has provided an interface 
using the Ethernet adapter on the motherboard. Similar interfaces 
using Belkin and Linksys USB-Ethernet adapters.  These interfaces are 
reliable.  They always work with the names assigned by udev.

> Whenever it's aesthetic is in the eye of the beholder.

Accepted.  Will happily give up aesthetics.  Whatever the name, 
the device should work.

> They give you pulseaudio for that.

OK, thanks.  Will try it with fingers crossed.  =8~)

I enjoyed reading https://wiki.debian.org/NetworkInterfaceNames .
Partly for the history. Logically, device identity should be settled 
at the system level rather than application level?  Hence 
systemd.link?  I guess it's just a matter of time.

> sox <infile> -t pulseaudio
> 
> And, since you're talking WAV - paplay <infile> .

Will try those also.  With any luck, later this week.

Thanks!            ... P.

-- 
https://en.wikibooks.org/wiki/Medical_Machines
Tel: +1 604 670 0140            Bcc: peter at easthope. ca


Reply to: