Re: Generalised approach to storing address details
Date: 9 Dec 2006 01:03:54 -0800
Neo a écrit :
> > > Cimode: ... EAV is pure crap....
> > If so, do you have any other suggestions, especially bearing in mind
> > Bernard's remark about international requirements?
> We can try a little experiment. If you would post a couple of typical
> addresses with different structures / order of fields, I can show how
> dbd's EAV-like approach would manage it. Others (Cimode?) can show
> non-EAV solutions. Then after the initial solutions are presented, you
> can specify additional new/different address structures that were
> initially unknown to see how each solution meets new data requirements.
> Neither approach is perfect. Each has it's advantages and
> disadvantages. If the ability to meet new/unknown data requirements is
> high on your list of importance, then your design will probably have
> some aspects of EAV and pay the price in terms of reduced dbms enforced
> constraints/performance and increased complexity of SQL/code.
A debate is interesting to launch when there is usefulness in doing so...EAV has been covered and in has been establlished that it breaks the Information Principle. Therefore it is bound to be under optimized in any circumstances...Therefore it is a total waste of time to discuss that subject... Received on Sat Dec 09 2006 - 10:03:54 CET