I've recently observed field names and data types are different.
1) For example, column names now use field name aliases, rather than field names.
2) And many fields, previously read in as "object" (using Pandas), are now defaulting to float64, int64.
#AGOL #shapefile