Sure I guess they were small but their problem space seems to be similar to a more standard CRUD. I could be wrong but the one to many and many to one relations

Sure I guess they were small but their problem space seems to be similar to a more standard CRUD. I could be wrong but the one to many and many to one relations within Twitter seem much harder (at least from my perspective). Both are super tough at the scale they run :)