Re: Like to implement GR_Image * m_pBackingImage in FV_View

On 5/2/07, msevior@physics.unimelb.edu.au
<msevior@physics.unimelb.edu.au> wrote:
>
> Hi everyone,
> I'd like to implement a new member variable for FV_View. It's
> a GR_Image * m_pBackingImage. It's a pointer to an image of
> the currently drawn page. FV_View will own the image and is
> responsible for deleting it.
>
> The idea is that if m_pBackingImage is ever non-NULL it is drawn wheneveer
> the FV_View::Draw(..) methods are called. This will enable the regular
> expose methods to work if we are doing fancy page animations. Expose
> events are generated from context menu's during a presentation. In this
> case a clip is set over the expose area and the image in the rectangle is
> blitted to the screen.
>
> We may also use m_pBackingImage for visual drag of images and text instead
> of doing a page redraw. It should be substantially faster and so give
> better results. However the main use and immediate use is to handle expose
> events from whole page drawings.
>
> What people's opinions?

Hi Martin,

not sure I understand correctly, but maybe it would be possible to
create a "save-under" on demand using gdk_pixbuf_get_from_drawable()
[1] and win32/osx counterparts?
This is a technique also employed by xservers (depending on
configuration), AFAIK.

For the VDND case the "save-under" might be the whole visible area,
would that work out?

[1] http://developer.gnome.org/doc/API/2.0/gdk/gdk-Pixbufs.html#gdk-pixbuf-get-from-drawable

- Rob
Received on Wed May 2 17:31:47 2007