Subj : Re: WM_PAINT or TWindow::Paint
To   : borland.public.cpp.borlandcpp
From : MarvinAvery
Date : Tue Aug 09 2005 05:09 pm


"Thomas Maeder [TeamB]" <maeder@glue.ch> wrote in message
news:m2iryfaucj.fsf@madbox2.local...

> > Thus, when calling dc.TextOut() even though I am writing the entire
> > line, only the last character actually gets written (or painted).
>
> Which is typically more efficient than writing the whole line.

Yes, I know.


> > Is there a way of telling or forcing Windows, for specific lines
> > only, to actually paint the *entire* line?
>
> You should be able to use InvalidateRect() for the rectangles occupied
> by these lines.

I can't use InvalidateRect() because the method doing the invalidating
doesn't know about these special lines.

>
> BUT: I don't see the point of your question. What's the difference of
> painting N-1 characters over their own graphical representation?

I am doing horizontal scrolling. My Paint() routine uses the scroll
positions to ascertain the characters to be painted. That is, it paints
characters, say, n to n+10. However some special lines are not to be
scrolled horizontally, thus characters 0 to 10 are always painted. When
scrolling, say, 1 character left, the *whole* window gets moved and only the
leftmost character gets painted. Thus, the special lines get scrolled; I
want to paint the entire line to "unscroll" it. I know I could use
Invalidate() but I would rather use UpdateWindow() as it looks much nicer in
action.

Now, to change the subject a little. Maybe you can help me here. I am *very*
confused about the following:
What is a device context's clipping region? What is it's update region? How
do these relate to the TRect that is passed as a parameter to
TWindow::Paint()?

TIA
Marvin

.