Andre Poenitz wrote:
> On Fri, May 18, 2007 at 08:43:12AM +0200, Peter Kümmel wrote:
>> Andre Poenitz wrote:
>>>> ===================================================================
>>>> --- src/frontends/qt4/GuiWorkArea.cpp      (revision 18380)
>>>> +++ src/frontends/qt4/GuiWorkArea.cpp      (working copy)
>>>> @@ -161,7 +161,7 @@
>>>>  
>>>>  GuiWorkArea::GuiWorkArea(int w, int h, int id, LyXView & lyx_view)
>>>>    : WorkArea(id, lyx_view), need_resize_(false), schedule_redraw_(false),
>>>> -    preedit_lines_(1)
>>>> +    preedit_lines_(1), delayed_scrollbar_sync(*new QTimer(this))
>>> Do you need an explicit timer object? There's QObject::startTimer()
>>> and QObject::timerEvent()...
>> Ah, didn't know there is already a timer when being a QObject. Will
>> have a look at it.
>>
>>
>>>> Index: src/frontends/qt4/GuiWorkArea.h
>>>> ===================================================================
>>>> --- src/frontends/qt4/GuiWorkArea.h        (revision 18380)
>>>> +++ src/frontends/qt4/GuiWorkArea.h        (working copy)
>>>> @@ -156,6 +156,14 @@
>>>>    /// timer to limit triple clicks
>>>>    void doubleClickTimeout();
>>>>  
>>>> +private Q_SLOTS:
>>>> +  /// process vertical scroll bar event
>>>> +  void verticalScrollBarActionTriggered(int action);
>>>> +  ///
>>>> +  void verticalScrollBarSliderPressed();
>>>> +  ///
>>>> +  void verticalScrollBarSliderReleased();
>>>> +
>>>>  private:
>>>>    /// The slot connected to SyntheticMouseEvent::timeout.
>>>>    void generateSyntheticMouseEvent();
>>>> @@ -177,6 +185,8 @@
>>>>    bool schedule_redraw_;
>>>>    ///
>>>>    int preedit_lines_;
>>>> +  ///
>>>> +  QTimer& delayed_scrollbar_sync;
>>> A _reference_? Why?
>> At could forward declare QTimer, but isn't that important.
> 
> This would be possible with a pointer, to.

But then you are not forced by the compiler to initialize it,
because of this a prefer a reference, it's shorter in declaring
and using than *const.

> 
> What you wrote works indeed, but I am not sure that it is
> clear to everybody glancing over the code _why_ it works.
> 

I'm preparing a patch which works, and where I know why.

> Andre'
> 

-- 
Peter Kümmel

Reply via email to