The problem is that after a discont, set_min_frame_size(1024) is called when
detect_stream returns FALSE. However, detect_stream calls check_adts_frame
which sets the frame size on its own to something larger than 1024. This is the
same situation as in the beginning, so the base class ends up calling
check_valid_frame in an endless loop.
The whole raison d'etre of this typefinder is to help avoid false
positives when used in combination with our main typefinder in
-base. Its task is not really to typefind audio/video files, even
less so given that it detects *MIME* types, not GStreamer media
types. Therefore, if this typefinder detects an audio or video
type, don't even suggest it - our own typefinders are hopefully
better at this and detect the right type of type on top of that,
and even if they're not we really want to know about it and fix
it. Fixes#575157.
Application developers won't know right away which module an interface comes from,
and may assume that it is covered by the usual GStreamer API guarantees, so make
it as clear as possible that this particular API is still subject to change
(should have done that with other libraries in -bad before too really).
If you're running things uninstalled, you might need to manually
remove the registry to get the plugin's features to show up
correctly. Also, you probably want to do a make clean to get rid
of the old plugin file.
fsselector is too close to selector, so import is as autoconvert. Not importing
the history because the directory is also called gst/selector and it fails.
Fix compiler warning due to condition always being true - since we're
only looking at the final byte of the startcode (not the leading
0x000001), only check for < 0x1f, not < 0x11f
We only need a Video Object Start code before we can start pushing out data.
Search for this code also instead of only looking for VOS and VOP.
Fixes#572551.
Baseparse internaly breaks the semantics of a _chain function by calling it with
buffer==NULL. The reson I belived it was okay to remove it was that there is
also an unchecked access to buffer later in _chain. Actually that code is wrong,
as it most probably wants to set discont on the outgoing buffer.
No short-desc as we have them in the element details.
Also keep things (Makefile.am and sections.txt) sorted.
Reword ambigous returns. No text after since please.
don't ignore the return value of pull_range because we only get a valid non-NULL
buffer when the return value is GST_FLOW_OK. Avoids a crash when the pipeline is
shutting down.
