I just applied the patch and can confirm that two finger scrolling works
famously.
My apologies for the keybinding. Instead of option, I meant to say "command".
However, the behavior is still different than other apple programs. In the
most recent SVN of LyX, command + right arrow advances a single word. The
typical behavior, according to the Apple support page, is for Command + Right
Arrow to move the text insertion point to the end of the current line.
Currently, that behavior is performed by Function + Right Arrow. This
distinction doesn't bother me (as I find the ability to navigate between words
valuable), but it deviates from the expected platform behavior and is therefore
worth mentioning. (I haven't looked at the code yet to see if it is related to
the other issue Pavel asked about.)
Cheers,
Rob
On Jun 30, 2010, at 6:45 AM, James Sutherland wrote:
> On Jun 30, 2010, at 4:56 AM, Jean-Marc LASGOUTTES <lasgout...@lyx.org> wrote:
>
>> "James C. Sutherland" <james.sutherl...@utah.edu> writes:
>>> When trying to scroll with the mouse, I get many copies of the following
>>> message:
>>>
>>>> /Users/stephan/cvs/lyx/lyx-2.0.0alpha4/src/frontends/qt4/GuiWorkArea.cpp(840):
>>>> wheelScrollLines = 3 delta = 0 scroll_value = 0 page_step = 734
>>>>
>>
>> The problem is "delata = 0". The code reads:
>>
>> void GuiWorkArea::wheelEvent(QWheelEvent * ev)
>> {
>> // Wheel rotation by one notch results in a delta() of 120 (see
>> // documentation of QWheelEvent)
>> int const delta = ev->delta() / 120;
>>
>> On PCs, the mouse wheel clicks, and each wheelEvent corresponds to a
>> multiple of 120 (unot is eighth of degree, and a click is 15 degrees).
>> On a mac, the situation is different, since the wheel turns smoothly.
>>
>> It might be enough to have delta be a double. Could you test the
>> attached patch?
>>
>> JMarc
>>
>> <scroll.diff>
>
>
> I am not really set up to build LyX from source right now, but would
> be happy to test a binary.
>
> Perhaps Rob is willing to build from source? Or maybe Uwe is willing
> to send out a binary?
>
> James
