On 08/12/15 11:22, Hendrik Leppkes wrote: > On Tue, Dec 8, 2015 at 11:14 AM, Luca Barbato <[email protected]> wrote: >> On 08/12/15 08:42, Martin Storsjö wrote: >>> From: Rodger Combs <[email protected]> >>> >>> --- >>> doc/APIchanges | 3 +++ >>> libavutil/opt.c | 12 ++++++++++-- >>> libavutil/opt.h | 10 ++++++++++ >>> libavutil/version.h | 2 +- >>> 4 files changed, 24 insertions(+), 3 deletions(-) >>> >>> diff --git a/doc/APIchanges b/doc/APIchanges >>> index 4c5e32e..bbffbb4 100644 >>> --- a/doc/APIchanges >>> +++ b/doc/APIchanges >>> @@ -13,6 +13,9 @@ libavutil: 2015-08-28 >>> >>> API changes, most recent first: >>> >>> +2015-xx-xx - xxxxxxx - lavu 55.4.0 - opt.h >>> + Add AV_OPT_ALLOW_NULL >> >> Actually, wouldn't be simpler to return a not-found error? >> > > But the option exists, its just set to NULL. Returning an not-found > would be rather confusing, as you couldn't differentiate.
My problem is that it gets another special case, while you'd like to query if the option exists or the option is set. So I'd use a flag to differentiate if we do not want to have two different errors for not found, but I'd rather not have something special just for 2 types and not the others. lu _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
