> This is not complete. V4L2_FIELD_INTERLACED means what? top-field-first? > bottom-field-first? See Michaels comments ... > > * drop open flags > > obsolete > > I remember some discussion that we could support O_NONBLOCK to make > capturing (syncing on a buffer, VIDEO_DQBUF) blocking or nonblocking > depending on this flag... Is that idea still pending, or is it discarded > as well? I still like the idea to have VIDEO_DQBUF look at O_NONBLOCK and then either block or return -EAGAIN like read() does. > > what about the video output stuff btw. Does anybody use that? > > Yes, the zoran stuff uses it for output... There's one problem, though, > the 'capture' formats (MJPEG and several uncompressed formats) aren't > the same as the playback formats (MJPEG only), and this isn't > programmeable in v4l2 (yet)... Could an extra flag for this be added in > the format enumeration struct? Done. Also created a new "enum v4l2_buf_type" and renamed VIDIOC_ENUM_PIXFMT to VIDIOC_ENUM_FMT. Deleted VIDIOC_ENUM_FBUFFMT, you can do that using VIDIOC_ENUM_FMT + the new type field now. Gerd -- You can't please everybody. And usually if you _try_ to please everybody, the end result is one big mess. -- Linus Torvalds, 2002-04-20
