GitLab

Clicking on it brings the adjustable slider.
DoubleClicking doesn't do anything anymore.

This way the behaviour is more usable and less complex <joke>(don't be too KDEish!)</joke>

this patch features :
- persistent resizable main windows
- video confined to preexisting window size
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

commit. Thou I think it should invalidate current album art, so core would do
new search for it.

 Input manager now emits signal for album art and gives
 url to art itself, so CoverArt or Background-widgets don't
 need to parse it themself, and CoverArt don't need to use
 those callbacks. This ways both get same image too.

This doesn't work yet, however, since p_input == NULL in the input_manager at the time this is displayed...
A fix is needed.

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Qt: Change the rate scale in order to have 0.96x in the slider in order to help people who want to fix NTSC/PAL issues.

Also remove lots of includes.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

DOesn't show anything yet.

Even lenny has it... (etch doesn't)
If you want to maintain a Qt4.2 version, you are welcome.

The vout code invokes the Qt4 video widget through the Qt4 window
submodule. Hence, the video widget cannot call the vout window code.
This is where the embedded video crashes seem to come from.
Unfortunately, this commit trades the deadlock (on 1.0) or the crash
(on 0.9) for an assertion failure.

Introduction of an abstractController, that can:
- create most of the interface controller widgets
- execute their actions

All buttons connect themselve to a mapper and a doAction( id_action )
that centralize everything. The Action could go into a mainActionner, I guess. Opinions?

This cleans a lot the signal and exchanges between MI and Controller,
between Controller and FSC. The buttons do their own cooking and connect
directly to THEMIM or to some of the few signals of the Controller
(inputExist, inputHasVideo, inputCanRecord)

This reworks most of the HACKS of Teletext Buttons and AtoB Buttons

The FSC inherit from AbstractController and not Controller, which remove the b_fscreation HACK.

There will be some regressions, I tried my best to minimize them.

The code is generic enough to be able to customize the toolbars now. HAVE FUN!

Well, of course, X11/DirectX being so easy, there are n-1/n changes of being on the wrong screen (if you have n screen).
But at least, you can move and it will remember the position.
(cherry picked from commit 29b82ab8d1fb740b4be0d6a0d142a107ffe7be19)
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Code simplification and TimeLabel in the FSC.

PFff :D
(cherry picked from commit 65c2a9bf053f82e4a84192f530f042209a449cca)
(cherry picked from commit 834bd87073e19b49dca9d1ea608541e5885fb3cf)
(cherry picked from commit 6a02accb4ca190a3e2b2690f2130ac6a09d78a32)
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

The window is always created with a pre-existing video output,
and destroyed before the video outout.

Besides, using object IDs is not safe.

Moreover, use a function instead of a macro (like in 0.9-branch)
Partially cherry picked from commit 73ab1cd94773956e463009088bf5f477cec1ae80

We need to reimplement an on-the-fly option to record streams.
I know this commit is crappy...

Qt: Add a parent to the CoverArtLabel We should add parents to all QWidgets created, otherwise the widgets are leaked This commit is buggy (can leak input_items), because the whole Qt implementation is leaking QWidgets