Re: Hiring the Rowing-Forward 30%
Posted: Apr 18, 2008 7:10 AM
> > Dr. Codd (and others) realized that the data was what
> > counted, and so devised/discovered a way to build
> > from the data outward.
> There are also cases where making the data "King" is
> equally flawed. Mandating that all data be represented in
> tables is flawed too. No doubt there are domains where it
> works and works well, but there are others where it does
> not do so well.
Do you know cases where it has been tried?
I don't know of any cases for 30 years where the full relational model has been used, so I find it hard to say it's failed - I have hopes for it and have a halfway implementation helping with the systems I build (abstracting away most use of SQL) - but - it's incomplete and we still don't use it everywhere, so I can't even say "Here's a domain where it works".
So, if you know of cases where it's been tried, I'm interested. Not just using an SQL database - using relational throughout, including full relational support (ability to create and modify relations throughout the code) and using this for everything.
I would guess the source for such experience would be some kind of 4GL language, maybe one the myriad variants that was running around in the 80s, or possibly something done in Smalltalk?
And, if you've seen it fail, I'd also like to know what form of representation of hierarchies was used in the relational model, what made things fail, and whether the hierarchy representation was related to the failure.
I've been playing with this area, trying to find good representations and using them throughout the systems we build, and going fully relational is the most promising direction I've found, which is why I'm so keen on knowing about whatever failures have come up.