[GRASS-SVN] r33938 - grass-addons/raster/r.watershed2

[svn_grass at osgeo.org]

Author: mmetz
Date: 2008-10-20 09:20:34 -0400 (Mon, 20 Oct 2008)
New Revision: 33938

Modified:
   grass-addons/raster/r.watershed2/README
Log:
new README

Modified: grass-addons/raster/r.watershed2/README
===================================================================
--- grass-addons/raster/r.watershed2/README	2008-10-20 13:18:23 UTC (rev 33937)
+++ grass-addons/raster/r.watershed2/README	2008-10-20 13:20:34 UTC (rev 33938)
@@ -10,16 +10,16 @@
   and that the configured source tree is available.
   To compile and install the test version of r.watershed for GRASS,
   
-  1) copy the directory <r.watershed.fast> into the directory 
+  1) copy the directory <r.watershed2> into the directory 
      <raster> of the GRASS souorce tree
   
   2) edit <Makefile> in directory <raster>
      add the line
-     r.watershed.fast \
+     r.watershed2 \
      e.g. after 
      r.watershed \
     
-  3) change to directory <r.watershed.fast>
+  3) change to directory <r.watershed2>
      run make
     
   4) optionally change to GRASS source directory, something like 
@@ -30,16 +30,17 @@
   Testing
   =======
   
-  The new test version can be invoked with r.watershed.fast.
+  The new test version can be invoked with r.watershed2.
   The new sorting algorithm is implemented for both the ram and the
   segmented version.
   The new segmented version is still much slower than the new ram 
   version, but the new segmented version is already much faster than 
-  the original ram version. At least 1GB of RAM are recommended.
+  the original ram version. A new option is available for the segmented 
+  mode specifying the maximum amount of memory to be used in MB.
   Set up a watershed analysis with the original version, run it.
   The command will have been r.watershed <options>
   To test, use the same input options, but different output maps 
-  and run r.watershed.fast <test options>
+  and run r.watershed2 <test options>
   There should be no differences in the outputs of the original and 
   the new version.
   If you find critical differences between the original and the fast version,