Well, I'm not a fan of SQL. To me the math underpinnings are not well exposed, probably because the language doesn't "see" obvious relations and that because of

Well, I'm not a fan of SQL. To me the math underpinnings are not well exposed, probably because the language doesn't "see" obvious relations and that because of this, joins must be performed explicitly, over and over again, and come to dominate the query. Isn't it obvious that when a column has the name of a table that the idea is that the columns of that refactored out tables become available as if it were a part of the main table? When the query works regardless of the structure of the table, the database can be refactored with ease. Most such queries would simply specify column names, and a filter to apply on the records. Any joins needed, and which result from the structure of the database, would be inferred. Such a "NoJoinDB" would clearly boost productivity, and lots of applications could be written with no explicit join at all. A language like SQL seems to hide the simplicity of most queries! Please comment about the validity of this perspective.