[FFmpeg-devel] FFmpeg release 6.1 (SDR Plans)

[Rémi Denis-Courmont]

Le lauantaina 23. syyskuuta 2023, 9.49.47 EEST Neal Gompa a écrit :
> On Fri, Sep 22, 2023 at 12:33 PM Michael Niedermayer
> 
> <michael at niedermayer.cc> wrote:
> > > What does this mean? Does this mean an FFmpeg release containing code
> > > that
> > > interfaces with your SDR library? Or does it mean the library fully
> > > integrated into FFmpeg?
> > 
> > It depends on the code at the time of release. ATM there is no seperate
> > library, just a SDR input device.
> > creating a separte library and the related API/ABI needs to be done with
> > thought and care not something to rush quickly.
> 
> What does this code *do*? All these arguments about the SDR code, and
> I haven't seen it myself (because the email patch workflow really
> makes it hard to track this stuff down, and I assume it was submitted
> on list somewhere?)
> 
> If it's just taking SDR devices as inputs and allowing you to encode
> audio and video streams, I'm not sure why you *wouldn't* have this as
> something in libavdevice or extended from it as a separate library.

I'm pretty damn sure why: that's not how you normally deal with analog (or 
PCM) audio inputs and outputs. There *are* already a standard interfaces for 
the purpose: ALSA is the most common (others include JACK and Pipewire). It is 
already supported by FFmpeg - and hundreds of other open-source applications 
that would benefit from it too. (And yes, you *can* implement ALSA devices in 
userspace.)

There are simply no particular reasons why this should be tied to FFmpeg, and 
there are conversely no reasons to bloat the FFmpeg code with SDR code that 
will only be usable to one in a million with suitable radio equipment.

But really the main concern was the slippery slope argument: SDR code for FM 
may be small, but SDR code, say, DVB wouldn't be, and the line has to be drawn 
somewhere.

-- 
レミ・デニ-クールモン
http://www.remlab.net/