[FFmpeg-devel] [PATCH] lavu: header and documentation for AVWriter

[Michael Niedermayer]

On Wed, Sep 07, 2022 at 03:30:09PM +0200, Nicolas George wrote:
> Anton Khirnov (12022-09-02):
> > As I already said to you in private, I do not think the motivation and
> > use cases for this have been sufficiently established.
> > 
> > You claim this will bring massive advantages all over the place. You
> > should support these claims with some actual patches that demonstrate
> > these advantages on some real code.
> 
> I have not claimed “massive advantages all over the place”, I have
> claimed minor advantages all over the places and massive advantages for
> future API.
> 
> With that correction stated, I have already answered you that your
> inability to see the benefits is only proof that your work on FFmpeg is
> on areas that rarely need strings. In fact, if I remember correctly, the
> last time it happened you used the worst possible implementation and we
> had to be multiple persons insisting to get it fixed.
> 
> Therefore, I consider your objection to be orders of magnitude less
> important than the positive feedback I finally had got from people who
> do use strings routinely, who in particular have used BPrint, are
> familiar with the benefits it brought but also the limitations I want to
> fix.
> 
> If there are no more remarks in the next few days I will consider this
> approved and start polishing the implementations.

My oppinion about AVWriter was asked for, iam not sure my oppinion is
worth that much but ok.

1. I think the escalating tone of the thread is bad for peoples blood
pressure and as well for the eventual technical outcome as it pushed
decissions and support in an emotional direction not a direction fouded
in facts and solid choice.

2. If AVWriter was called bprint 2.0 you might have avoided this whole
opposition. And it really is IMHO a next generation bprint API.

3. You are the author of bprint and this is a from my point of view a
2nd generation bprint. I see nothing controversal on this from a technical
point of view. Its simply a step forward initiated by the original author
of the old code
I think we should move to this bprint 2.0 unless someone has a solid
argument why bprint 2.0 is worse than bprint 1.0. In which case maybe
that issue can then be fixed

thx


[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Nations do behave wisely once they have exhausted all other alternatives. 
-- Abba Eban
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: not available
URL: <https://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20220914/54ece6eb/attachment.sig>