I am busy setting up my layers for a typical vector map display, but I have a question around the use of definition queries instead of using split up layers.
I have a road network layer that contains a type field and obviously different types are displayed on different zoom scales. Now I have the option of using the definition query to load the roads of a specific type from the dataset, but the implication of doing this, is that on each different zoom scale, Arc requires all road types, which does take a while to render.
My question is would it be advisable to rather split out the layers into the different road types? The problem with this is that I won�??t be able to make use of the "join/merge" functionality with the Symbol Level? or Is there a way to load the full dataset once in memory, then use the same data on the different zoom scales (almost like a layer memory cache) thus limiting the cycles needed to query the raw vector data?