Lambda the Ultimate

In this Interview, Date discusses his new book, Database in Depth: Relational Theory for Practitioners. The book does look interesting and is on my list to be acquired at some point in the future. The Third Manifesto crowd can be a bit bellicose at times (at least for my tastes), but I can't say that I disagree with many of their assessments.

The underlying model (relational in this case) is of interest but more importantly is the manner in which it manifests itself within the language chosen for communication. SQL has some advantages as a language that I think extend beyond the realm of being ubiquitous. It primarily serves as a bridge between the programming languages that we choose, and the databases that we are stuck with. One could naturally ask the question of why have a seperate language in the first place (Object Databases tend to go this route)? And why do we need either SQL or D, when we have much more universal programming languages available at our disposal? If one surmizes that a database language is needed but that it should somehow be crippled in expressiveness (e.g. not turing complete, not imperative, not....), then one is forced to answer the question of where to draw the line.

That said, what I am currently wondering is if there is an implementation of the D language besides Dataphor? I guess I'm spoiled by the myriad of unencumbered programming languages that are being actively developed and I sometimes think that there is a cultural divide between those who implement programming languages and those that work with languages in the database market. Perhaps what we need is for an unassuming brilliance to appear out of seemingly nowhere and surprise us with an implementation of the language in Haskell or Scheme. Any more Autrijus' out there? :-)