ArcView Feature-creation Hanging

Question

When creating polygon features in ArcView 10, using the auto-complete polygon tool, I create a feature neighboring another polygon and ArcView hangs or processes for over 15 seconds. The feature is successfully created, but it takes forever. The polygon layer being edited is in a file geodatabase (although location of the feature, in or out of a gdb, being edited appears to have no bearing on the problem). I've done the Normal.mxt rename and all the tried-and-true solutions and ESRI has been of little help.

Any help would be appreciated, thanks. -Ross

Answer 1

Check the tolerance of the feature class: ArcCatalog > Right-click the feature class > Properties > Tolerance. A proper value depends on the spatial reference (and the associated linear unit) and what kind of precision you require.

For almost all purposes the default tolerance chosen by ArcGIS is good enough. If the tolerance is set too low (i.e. extremely fine-grained tolerance ) this results in large amounts of extra processing when interacting with the features but not when simply drawing them. As a point of comparison: for data in geographic WGS84 the default tolerance is 0.000000008983153 decimal degrees.

The tolerance of a feature class can only be set at the time of creation. If the tolerance value is your problem, create a new feature class with the proper settings and load the features into this new feature class.

Answer 2

This a response from ESRI Support (USA):

There are a few known issues that cause slow performance with version 10. Large datasets can certainly slow things down. Editing over a network, or using .mxd files that were created in a previous version are also known to degrade performance.

How large are the datasets that you are working on (size of file and number of features)? Line or polygon features with huge numbers of vertices, so called Godzillas, can also be a problem for performance. These can be diced into smaller features with fewer segments to improve processing time, but in general, large datasets require more processing power.