Re: gEDA-user: PCB+GL Progress

[Peter Clifton <pcjc2@xxxxxxxxx>]

On Mon, 2009-02-23 at 15:35 +0000, Kai-Martin Knaak wrote:
> On Mon, 23 Feb 2009 03:43:33 +0000, Kai-Martin Knaak wrote:
> 
> > First impressions with one of my more clouded layouts:
> >  
> > Thin draw of polygons looks nice. However, it is about as slow as full
> > polygons.

It _should_ be slower ;)

With the extra translucent fill in the thin-draw mode, you're not only
filling the polygons, but drawing lines around their edges. (Although
that later operation isn't particularly expensive now I fixed the stupid
case where I tried to draw circular caps on hairlines).

I suspect I could do better still if / when I add a list of GL_LINE
primitives to emit for "0px" lines, rather than making my "hardware"
lines out of two triangles, computed to give a width of 1px on screen.

>  See the redraw rates below. If this is added permanently, I
> > would ask for a single click option to completely disable polygon
> > display. Currently, my work-around is to put all polygons in special
> > layers and temporarily attach them to their own layer group.
> > 
> > Results of the benchmark() command on my not quite up-to-date desktop
> > hardware (3600 BogoMIPS, glxgears ~1500):
> > 
> > full polygons: 2.7 redraws/sec
> > thin draw polygons: 2.6 redraws/sec
> > polygon layers off: 126 redraws/sec
> 
> Redraw rate on my two years old desktop at my dayjob (4400 BogoMIPS, 
> glxgears ~890, nvidia quadro NV37GL):
> 
> full polygons: 3.5 redraws/sec
> thin draw polygons: 3.9 redraws/sec

If this is the code-base which has the translucent filled thin-draw
polygons, I'm surprised this is faster. It is probably just a
measurement error due to statistical variance. PCB's benchmark timer
isn't that accurate, it relies on scheduling latency, and only counts
the number of frames in a 10 second period. At 3.5fps, that isn't many
frames to average over.

Of course.. thin-draw (rather than just thin-draw poly) may well be
faster due to the simplification of other primitives.

> polygon layers off: 34.5 redraws/sec
> 
> 
> My print command segfaults with the GL enabled version:

Should now be fixed.

-- 
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-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user