Here a picture of methods:
http://www.rscheata.net/images/impso....methods_2.jpg

neXus is now over 20k lines of code.
excluding scripts, randoms, epascal, jyhton

Here is a 46mb video of some of those things above being used:
http://www.rscheata.net/videos/unitTest.htm

We are now working on making advance walking methods.

Object finding methods are now very very advance. Example:

This will find a bank booth, click it and select from menu to open quickly:
doActionAtRGBCluster(new RGB(70, 50, 10), 15, "Bank", "quickly");

The way it operates is:
1. divides the screen into zones and finds zones that have the highest density of the color.
2. It then sorts the zones by the density.
3. Removes zones that are in proximity to each other.
4. Cycles thru results looking for topText in each zone.
5. if correct top text, open menu and select the text, if text selected return
6. repeat 1-5 till for set period of time till success

Also: zones overlap so that no results are affected by the borders of zones.

Now in testing this method it could see that sometimes zone existed that were not my goal.

To solve this I added in using multiple colors, Also the preference of first testing zones that had all the colors while still using some zones that had partial matches for fail-safe.

doActionAtRGBsCluster(new RGB[]{new RGB(70, 50, 10),new RGB(136, 107, 46)}, 15, "Bank", "quickly");