On Thu, Mar 3, 2022 at 1:23 PM Richard Purdie
<[email protected]> wrote:
>
> On Thu, 2022-03-03 at 12:54 +0200, Alexandru Ardelean wrote:
> > It didn't look [to me] like it would be too bad to me either.
> > I wasn't sure what the decision would have been (from OE), to
> > [1]patch-out libopus? or to [2]add libopus? or [3]revert?
> > But your patch suggests [1] is preferred for now.
>
> My reasoning is that:
>
> * libopus is in meta-openembedded, I'm not sure we have a strong case for 
> moving
> it to core.
> * I was happy to see us move to a recent more up to date version with the CVE
> patches dropped so reverting seemed like a less attractive option too.
>
> I just wasn't sure if patching out libopus would be possible or not. I 
> therefore
> had a look at a patch to do it and that shows it does seem to be reasonable.
>
> > Did you get a chance to test it?
> > I can run a test on my system with it.
> > But it may take me 1-2 days.
>
> I've turned it into a proper patch and will test it on the autobuilder.
>
> > In the meantime, I was a bit busy with this:
> > https://github.com/libsndfile/libsndfile/pull/812
> >
> > It kind of seems that libsndfile people are talking about maybe
> > splitting up their configuration.
> > Let's see. That will be in a future release anyway.
>
> Thanks for discussing it with them. If upstream were willing to accept a patch
> to enable/disable opus, that would be our much preferred way to handle this!

Tested your patch no-opus patch.
It works for me.
I have a running system and I'm also able to convert the OGG Vorbis
files to WAV.

Regarding libsndfile, things are a bit on hold.
I may bother them next week.
Reference: 
https://github.com/libsndfile/libsndfile/pull/812#issuecomment-1059762147

It looks like there was a proposal (3 years ago) to split the
configuration, but it was NACK'ed by the main maintainer/author of
libsndfile.
Reference: https://github.com/libsndfile/libsndfile/pull/400

Now, with (me) opening the discussion PR 812, it looks like this is
being re-evaluated again.

Good news is: there is already work done for this (splitting
configuration per Xiph-codec), it just needs to be pulled back into
the present.

Thanks again
Alex

>
> Cheers,
>
> Richard
>
>
>
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#162817): 
https://lists.openembedded.org/g/openembedded-core/message/162817
Mute This Topic: https://lists.openembedded.org/mt/89367260/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub 
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to