I will explain EAV with a very short example.
Table ANALYSIS
---------------------
IdAnalysis
Date
XCoor
YCoor
UTMHuse
Table RESULT
------------------
IdAnalysis
IdComponent
IdValue
Table COMPONENT
------------------------
IdComponent
ComponentDefinition
Unit
Where analysis owns several results, and the kind and number of components in every analysis varies.
EAV modelling becomes from deep normalizing low estructured data. At first, I was strong supportting EAV because opponent's solution was just no-normalization at all.
But as time went by, I realised useful views was neede in order to work with data, and to design forms for these tables was not very clean.
So EAV modelling is an example about how technical users or plain users can diverge in their models, and how they both should be equally satisfied.
That's something I would like to dive in in this research. We will see.
No hay comentarios:
Publicar un comentario