[GRASSLIST:4176] r.contour... Followup

[cbsled at ncia.net]

I reclassed the DEM to feet, then ran r.contour to generate the contour for the single problematic level, followed by v.out.ascii to v.out.dxf. The resulting contour does not exhibit the short segment problem. The segment count on the test section has dropped from 2278 to 1848. From those numbers, it should be obvious that manual fixing is out of the question.

I think this narrows it down. The problem appears to occur in special circumstances and only where the specified contour value is not an even multiple of the elevation resolution of the DEM. 

A floating point problem perhaps? 

I would be happy to send the offensive and inoffensive output files (total size<1MB), or any other useful information, to anyone interested. 

-- 
-----------------------------------------------------------
Carl Brown cbsled at ncia.net
------------------------------------------------------
-----