On Thu, Dec 29, 2016 at 03:33:54PM +0100, Nicolas George wrote:
> Signed-off-by: Nicolas George
> ---
> libavfilter/avfilter.c | 52
> ++
> libavfilter/filters.h | 42
> 2 files changed, 94 insertions(+)
>
Signed-off-by: Nicolas George
---
libavfilter/avfilter.c | 52 ++
libavfilter/filters.h | 42
2 files changed, 94 insertions(+)
Changes in this commit: rename ff_link -> ff_inlink and move to filters.h,
re
On Sun, Dec 25, 2016 at 11:44:07AM +0100, Nicolas George wrote:
> Le quintidi 5 nivôse, an CCXXV, Michael Niedermayer a écrit :
> > Whats your oppinion on using a explicit av_assert1() in the calling
> > code for this ? (i assume it can be done easily&cleanly)
> >
> > It would explicitly in C code
Le quintidi 5 nivôse, an CCXXV, Michael Niedermayer a écrit :
> Whats your oppinion on using a explicit av_assert1() in the calling
> code for this ? (i assume it can be done easily&cleanly)
>
> It would explicitly in C code say what is meant, while a
> "_sure" requires additional knowledge specif
On Sun, Dec 25, 2016 at 02:00:13AM +0100, Nicolas George wrote:
> Le quintidi 5 nivôse, an CCXXV, Michael Niedermayer a écrit :
> > Why do these functions exist twice ?
> > the API would be simpler with just one set
>
> You mean the _samples / _frame versions, or the normal / _sure versions?
i me
Le quintidi 5 nivôse, an CCXXV, Michael Niedermayer a écrit :
> Why do these functions exist twice ?
> the API would be simpler with just one set
You mean the _samples / _frame versions, or the normal / _sure versions?
The _samples / _frame case seemed easier than distinguishing using
special val
On Sat, Dec 24, 2016 at 06:41:40PM +0100, Nicolas George wrote:
> Signed-off-by: Nicolas George
> ---
> libavfilter/avfilter.c | 63
> ++
> libavfilter/internal.h | 47 +
> 2 files changed, 110 insertions(+)
[...
Signed-off-by: Nicolas George
---
libavfilter/avfilter.c | 63 ++
libavfilter/internal.h | 47 +
2 files changed, 110 insertions(+)
diff --git a/libavfilter/avfilter.c b/libavfilter/avfilter.c
index 22a1fedd17..e
