Rostislav Pehlivanov (12019-01-04):
> > +typedef struct AVRegionOfInterest {
> > + size_t self_size;
> > + size_t top;
> > + size_t bottom;
> > + size_t left;
> > + size_t right;
> I'd still much rather have uints with fixed sizes than these platform
> dependent types.Guo, Yejun said: >> I usually choose 'size_t' for the meanings with length/size. But that is a mistake. size_t is for length/size of objects in memory, not any length/size. These numbers, unless I am mistaken, are coordinates within an AVFrame. In that case, the only correct type is the same as AVFrame.width and AVFrame.height. > Use an AVRational with a denum set to the max quantizer Can you explain "set to the max quantizer"? For decoders it makes sense, but what should the encoders do? Return EINVAL? Scale? Regards, -- Nicolas George
signature.asc
Description: PGP signature
_______________________________________________ ffmpeg-devel mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
