Re: [PATCH 1/2] pcm: hw: reinterpret the drain_silence setting

[Oswald Buddenhagen <oswald.buddenhagen@gmx.de>]

On Mon, May 15, 2023 at 11:06:54AM +0200, Jaroslav Kysela wrote:
>On 15. 05. 23 10:41, Oswald Buddenhagen wrote:
>> It makes no sense to have a config option which determines the entire
>> silencing time, because useful values strongly depend on the period size
>> runtime setting.
>
>I would prefer a drain silence mode configuration field for this, if the 
>interpretation of the positive value is different. The current value is easy 
>to understand and it's really a corner case to not use the default (auto) 
>code.
>
but why would you *want* that? what would one use it for? what is the 
advantage over my proposal?
i don't think "easier to understand" is true. it may seem easier on the 
surface, but one cannot make an informed decision without actually 
understanding the implications.

>Even the period time may be forced in the configuration, so users can 
>compute the right value themselves.
>
that means that only by combining the two options you'd actually get 
something that makes sense.
my proposal otoh works both with and without a forced period time.

regards