Re: gEDA-dev: World vs. screen coordinates...

[Peter Clifton <pcjc2-at-NOSPAMcam_dot_ac_dot_uk>]

On Sat, 2006-12-23 at 20:16 +0100, Carlos Nieves Ónega wrote:
> Hi Peter,
> You are right. Screen coords and world coords have different sense.
> You can check it opening gschem, and showing the coordinates windows (oc
> hotkey).
> 
> Regards,
> 
> Carlos

So it is... I've never noticed the coordinate window before! I remember
having some difficulties with bounds checking for hit-detection in world
coords - (And grip locations too), due to the apparent need to change
the sign of any addition or subtraction intended to operate in a
particular direction.

I think then, there are a few bugs in gschem's object world bounds
functions. (I've corrected in my big diff which I'm separating).
Sometimes, the code assigns *top for a world bound, to the min() of
various coords, where for world coords it should be max() according to
our observations in the coord window.

Cases I've got fixed here are:

world_get_box_bounds()
world_get_bus_bounds()
world_get_circle_bounds()
world_get_line_bounds()
world_get_net_bounds()
world_get_picture_bounds()
world_get_pin_bounds()

(I've changed many implementations to look like:

+  *left = min( line->x[0], line->x[1] );
+  *top = max( line->y[0], line->y[1] );
+  *right = max( line->x[0], line->x[1] );
+  *bottom = min( line->y[0], line->y[1] );

)

This anomaly is quite prevalent thought the code.. Is it possible that
world coordinates were never _meant_ to be opposite to screen-coords in
up/down sense?

Ales - is there a definition of world / screen coords, and their origins
+ senses?


Happy Holidays... its 25th now, so Merry Christmas to those who
celebrate it!

Peter C.




_______________________________________________
geda-dev mailing list
geda-dev@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev