Re: Tragic State of FreeBSD Audio/Sound

On Wed Jul 22, 2026 at 10:07 AM EEST, Michael Gmelin wrote:
>
>
> On Wed, 22 Jul 2026 02:25:08 +0300
> "Christos Margiolis" <christos@freebsd.org> wrote:
>
>> 
>> I do not object to the fact that the change was invasive and broke
>> scripts, but I do disagree completely that the new syntax is more
>> confusing or worse. It can naturally support more functionality and is
>> structured in a well-known format: foo.bar=value.
>> 
>
> Just to add my perspective, I would have preferred to have this change
> introduced as something like "newmixer" or "mixer+", so the existing
> interface wouldn't change (could be the same binary checking argv[0] to
> determine behavior).

The point was to take the "hit" and create a cleaner/better interface
for mixer(8) once and for all, not introduce yet another tool and
clutter FreeBSD more.

> I'm the kind of person who regularly does things like
>
>   ssh host mixer vol -10
>
> and I was a bit disheartened when the decades old syntax/finger memory
> now returned
>
>   mixer: -10: no such device

You can still do:

	ssh host mixer vol=-10%
>
> (like: "great, now even adjusting the volume doesn't work like it used
> to anymore, I'm feeling old").
>
> I don't know how many people use mixer on the command line on a daily
> basis - fixing your desktop keybindings once is not a big deal, but
> forgetting and relearning syntax is - and I certainly wouldn't get out
> of my way to insult anyone over this change, but I was definitely
> astonished, so for me it certainly was an (avoidable) POLA violation.

It is a POLA violation, I don't disagree. I just do not think that
keeping broken or limited interfaces just for the sake of an arbitrary
notion of POLA is *always* better.

> Cheers
> Michael

Christos