> This mechanism, nicknamed "place oriented programming" by Rich Hickey [REF], saves storage space but makes it impossible to retrieve the entire history of a particular record. This design decision reflects the computer scientist’s perspective that ‘history’ is less important than the price of its storage.

Funny, I don't recall that sentiment from any of my CSCI classes. Seems to me like this methodology has more to do with the fact that a 5 MB hard disk (for example) used to be the size of a washing machine than it does with whether or not the person writing the DBMS was a computer scientist.

Yes absolutely. But the question is why is this methodology still prevailed even though we are not constrained anymore by these limits?

Don't these already exist?

http://www.datomic.com/

Seems to be an attempt at reimplementing at least parts of datomic as open source: https://github.com/aosabook/500lines/tree/master/functionalD...