On Mon, 1 Dec 2014 13:56:57 +0100 Nicolas George <[email protected]> wrote:
> Le primidi 11 frimaire, an CCXXIII, wm4 a écrit : > > I did not expect that it would crash. And why should it? > > Because it does not make any sense. I used it for making the decoder release all surfaces before uninitializing the hardware decoder. Or are you suggesting to add multiple cleanup code paths? > Next time, they will ask the C standard to accept dereferencing NULL, that > is the same logic. If you want to use the C standard as baseline: C11 introduced a bunch of functions like c32rtomb(). They accept a NULL parameter for the state, in which case global state is used. > If the application is not capable to know whether a context is opened or > closed, then there is something deeply wrong in its design. It's hard to write sane code when the FFmpeg APIs are so crappy. > Sprinkling FFmpeg's code to work with random tests to work around broken > application designs is not good for maintainability. Just look what's going on in FFmpeg development. Can you seriously claim maintainability is a goal? There are so many changes that should have been rejected for the sake of maintainability. (But this is offtopic.) _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
