[FFmpeg-devel] [PATCH] avcodec/imgconvert: remove deprecation guards for a function that's not declared as such

Ronald S. Bultje rsbultje at gmail.com
Fri Aug 25 17:39:38 EEST 2017


On Fri, Aug 25, 2017 at 10:04 AM, James Almer <jamrial at gmail.com> wrote:

> On 8/25/2017 8:10 AM, Ronald S. Bultje wrote:
> > Hi,
> >
> > On Thu, Aug 24, 2017 at 7:43 PM, James Almer <jamrial at gmail.com> wrote:
> >
> >> Signed-off-by: James Almer <jamrial at gmail.com>
> >> ---
> >> The deprecation seems to have been imported from libav but never made
> >> effective.
> >
> >
> > Hm, but do we really want this function? Shouldn't all users be ported to
> > the function this wraps?
> I don't know. The Doxy even acknowledges there's an alternative but
> mentions its use case is apparently different, which is probably why
> the deprecation wasn't made effective.

We should all acknowledge fully that there is a use case for
memcpy_inverted(source, destination, size), yet libc does not define it. I
don't know why. It's silly. I feel libc is being racist.

Silliness aside, let's not have multiple APIs that do the same thing.

If the function really needs to go, then the deprecated attribute should
> be added to the header. And from that point the ~2 years deprecation
> period starts.

Sure, operationally I don't really care, as long as the end product is that
it goes away.

Fork nastiness aside, I feel that one thing Libav does have a much better
handle on than us is API cleanliness.


More information about the ffmpeg-devel mailing list