The param option b_vfr_input is only available in x264 >= 81.

The mozilla webplugin supports only the libvlc_MediaPlayer* events just like the ActiveX webplugin.
Ommit the 'libvlc_' prefix from the libvlc_MediaPlayer* events to obtain the events name that must be
used from JavaScript to listen on the event. Example:

- libvlc_MediaPlayerOpening becomes 'MediaPlayerOpening'
- libvlc_MediaPlayerNothingSpecial becoms 'MediaPlayerNothingSpecial'

In the windows gecko-sdk the call NPN_PluginThreadAsyncCall() is not supported, so the code most resort to an older method of sending async events from the plugin to the browser.
Source: http://stackoverflow.com/questions/1931883/generating-async-javascript-events-from-browser-plugin-npapi

The mouse-release event signals that the video-filter detected a release on the object. This is needed,
because it is not possible to detect release from within JavaScript reliably and have the filter (eg: logo)
update its objects position during the dragging.

Do not show ":" separator if album missing.
Show album even if artist missing.
Optimize a little bit.

OS X framework: fixed packaging failure, which prevented Lunettes execution on some setups due to a failing libvlc initialization

especially, also detect material recorded at the stream filter level

same options needed for vlc_accept as those for vlc_socket

some video seemed to go pass and sended to last.fm otherwise.

Signed-off-by: Rafaël Carré <rafael.carre@gmail.com>

radio-france.fr seems to use such URLs
example: http://son.radio-france.fr/chaines/itema/10491/10491_02042010_2010F10491S0092.rm

All hail to our new owners !

MainInterface:
  Use same settings context when creating and deleting playlistWidget
  Make beginGroup() / endGroup() pairs explicit.
StandardPLPanel:
  Use same settings context whenever creating treeView

This is still blind fix, don't have windows box to test atm. So don't blame
me when it eats your cat and runs to riviera with your money and gf (and free
to revert in those cases). Don't remember if that check works correctly in
windows side.

I didn't observe any issues with 1.1 or 1.0.5 from file-input, but
lets just set it just in case when we have fps-value.

This should also fix the accept4() ENOSYS infinite loop in httpd
pointed out by Aurélion Nephtali.