Hi,

On Fri, Apr 8, 2011 at 10:15 AM, Anton Khirnov <[email protected]> wrote:
> On Fri, Apr 08, 2011 at 10:10:18AM -0400, Ronald S. Bultje wrote:
>> On Fri, Apr 8, 2011 at 9:43 AM, Anton Khirnov <[email protected]> wrote:
>> > On Fri, Apr 08, 2011 at 07:39:52AM -0400, Ronald S. Bultje wrote:
>> >> On Fri, Apr 8, 2011 at 5:48 AM, Anton Khirnov <[email protected]> wrote:
>> >> > ---
>> >> >  libavformat/avformat.h |    4 +++-
>> >> >  libavformat/utils.c    |    4 ++++
>> >> >  libavformat/version.h  |    3 +++
>> >> >  3 files changed, 10 insertions(+), 1 deletions(-)
>> >>
>> >> Any reason we don't want to remove it at the next one already, going
>> >> by the evil "release, bump, release again" plan?
>> >>
>> >
>> > As you already know =p, I think our policy should be to remove old stuff
>> > only after a sufficiently (at least ~0.5y) long transition period.
>>
>> These *2() functions have been out for a few months, so it's long
>> enough. I've keen on removing as much old cruft as possible.
>>
>> (It's not like this is critical API that every application definitely
>> uses, unlike say avcodec_decode_audioX() or so... The chance of real
>> problems is relatively small.)
>
> Only one month.
>
> Don't forget that we already have a bad reputation for breaking
> compatibility often (undeserved, but still...). I'd prefer to be more
> conservative with those things, unless they obstruct progress. That is
> certainly not the case here.

Fair enough, I already OK'ed the patch. But if the code was bigger or
it would obstruct reordering of stuff in structs (i.e. ABI) I'd make a
bigger deal about it maybe.

Ronald
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to