r.surf.contour slowdown

[Chuck Ehlschlaeger]

>r.surf.contour indicated that it finished about 10% of the job in the
>first hour, but here it is two days later and it indicates that it's
>only done 24% of the job -- only 1% progress last night.  Why has it 
>slowed down so much?  

The speed at which r.surf.contour runs depends on how far the 
contours are from each other at the region of the map being calculated.
If the "top" tenth of the map had closely packed contour lines while the
next section's lines were far apart, that could cause the problem.  If
you have large sections of areas w/out contours, this will cause significant
slowdowns.  Carefully read the last part of the r.surf.contour MAN page.
The MAN page contains many hints to optimize the program.

>I figured that running r.patch before r.surf.contour would do a smoother
>job along the edges of the quadrangles, but I suppose I could reverse
>the order of those operations.  Any insights? 

You are correct on the assumption of a smoother job.  Another way to do
this would be divide up the job into several regions, but overlap the 
regions by at least one contour interval.  Patch together the results using
the closer half of the overlap on each side.

chuck