* Sven Neumann ([EMAIL PROTECTED]) wrote:
> Hi,
> 
> Hallvar Helleseth <[EMAIL PROTECTED]> writes:
> 
> > I've been thinking about a new way of using StretchBlit.
> > I think StretchBlit could be used to stretch a surface in atleast four
> > ways:
> > 1. Tileing the source surface to fit the destination surface
> >    (remove the TileBlit function)
> > 2. Scaling up/down the source surface while keeping borders
> >    (think of Imlib2)
> > 3. Tileing the source surface while keeping borders.
> > 4. Just scale up/down, like it does today.
> > 
> > A way of doing this could be to add:
> > DSBLIT_TILE, DSBLIT_BORDER, DSBLIT_BORDER_TILE to DFBSurfaceBlittingFlags.
> > 
> > In IDirectFBSurface_StretchBlit() (display/idirectfbsurface.c) you could
> > check for the blit flags and call the appropriate
> > dfb_gfxcard_stretchblit(), dfb_gfxcard_tileblit(), or
> > dfb_gfxcard_borderblit()
> > 
> > For a surface to know where the border is I think we need a new API call:
> > surface->SetBorder(surface, left, right, top, bottom)
> > which would alter the (struct _CoreSurface).left_border/right_border/etc
> > Perhaps also the ability to specify the border upon surface creation?
> > 
> > The reason I want all these four functions in StretchBlit is because
> > they all stretch blit (part of) a source surface to (part of) a
> > desitination surface, just in four different ways...
> > 
> > Or maybe you'd like the four functions to have their own API call ?
> > (I really want the border blit function, which I've already
> > implemented at the application level, but it's much nicer to have in the
> > library since it's such a usefull function!)
> > 
> > Anyways if you give me thumbs up I'll be happy to implement it myself as
> > good as I can, and send patches.
> 
> IMO this feature belongs into the application level. It's rarely
> useful and can easily be implemented outside of DirectFB if it is
> needed. If you think it could be useful for other apps too, why don't
> you put it into a utility library so the functionality can be shared
> without bloating the DirectFB API? Please keep in mind that we care a
> lot about the size of the DirectFB library since our main target is
> embedded devices.

Makes sense. This I why I asked _before_ I made a patch. But I don't
want to bloat the mailing list with losts of stupid questions either,
seeing as you people seem pretty buisy ;)


Hallvar Helleseth


-- 
Info:  To unsubscribe send a mail to [EMAIL PROTECTED] with 
"unsubscribe directfb-dev" as subject.

Reply via email to