On Sun, 2008-06-22 at 19:09 +0000, Kai-Martin Knaak wrote:

> More notes on the openGL amusement accumulated when I tried to do sone 
> real work:

Thanks (most of this GL stuff was done after I had the majority of my
board made, so I haven't used it extensively).

> * The view window blanks at certain events:
>     --> If an overlapping window is moved in front of pcb --> When a tool
>     tip about a window decoration closes --> When a pcb menu closes
>     without selecting an action --> When a sub menu closes
>     --> When gnome shows the mouse locator (as mentioned before)
>   I'd guess, these are all symptoms of the same redraw illness.

Yes, I'll have a go at squashing those.

> * If the preference dialog closes, the exposed part of the viewport is
> rendered without transparency until the next redraw event.

Bizzare!

> * visible grid does not show.
> * The cross at the origin set by [CTRL-m] does not show.
> * The little L at the handle of locked elements does not show.

I need to re-code GL versions of those, shouldn't be too hard - perhaps
a few lines of code each.

> * The x coordinate of the cross hair does not snap to grid points. It only
> snaps to pins, pads and end of tracks. However, the y coordinate does snap
> to invisible grid points.

Sounds like a bug in my updated snapping behaviour (which is also
applied in that patch). Thanks for the detailed report.. I felt
"something" might have been wrong with it, but hadn't quite put my
finger on it.

> * There is no preview if I drag a track with the mouse.

Ok.. checking, something seems to have regressed there. I suspect I'm
generating degenerate triangles with coincident vertices, which GL is
not rendering. I need to fall back to rendering with lines for "thin"
objects. Oddly though, thin-draw appears to work!

> Unfortunately, the last two bugs make the GL-patch next to unusable for
> manual routing. Hopefully, they are easy to squash. Thanks again for 
> pushing geda ahead!

I'll work up an updated patch in the next couple of days. I already have
improved code for polygons, avoiding those memory leaks which added up
over time, and avoiding malloc for most polygons (at least for the
complexity of polygons in my current board).

Best wishes, and thanks!

-- 
Peter Clifton

Electrical Engineering Division,
Engineering Department,
University of Cambridge,
9, JJ Thomson Avenue,
Cambridge
CB3 0FA

Tel: +44 (0)7729 980173 - (No signal in the lab!)



_______________________________________________
geda-dev mailing list
[email protected]
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev

Reply via email to