To provide a little more information on my question, we are making choropleth maps covering the entire world. The data varies widely country-by-country, such that 5 classes at the global scale should have very different class breaks than 5 classes of a high producing country like Brazil or a low producing country like Cuba or Bangladesh. So online, the dataset would have, what I believe you call a dynamic legend, recognizing the change in the range of values depending upon the geography being viewed. True? Also, my question is what control could developers have on that dynamic legend? I don’t want class breaks of 534, 1248, 1783042 (as examples). Instead I want class breaks of 500, 1250, 1783000. Can we set the dynamic legend to always round to the nearest 100th value or 50th value?