oktober 22, 2008
Complete Rewrite: Intressant blogg om databaser och systemutveckling
Complete Rewrite är en ny intressant blogg om Personal viewpoints on database systems and computer programming.
Passionate programmer since I got my first programmable calculator in 1985, my work developing search engine and data management software, as well as my own understanding of database systems and software design, at Apptus Technologies has taken me quite some distance from my background as a practical algorithm theoretician with Ph.D. on suffix trees and data compression, and allowed me to develop insights and opinions on such diverse areas as object-oriented design, the relational data model, transactional systems, development processes, and logic.
Till dags dato har skrivits två (långa) artiklar:
There is hardly any field in computing more plagued by religious wars than database systems. For nearly 40 years, the battle has raged between various architechtures, occasionally with some combattants replaced – or at least renamed. Still, it seems that we are further away than ever from a sort of database that we can be satisfied with. (I am not going into the details of the problems right now, that will be a subject of future posts. But if you are in the business I am sure you are familiar with some of them.)
It is particularly common to write off the relational model in the text search field. You can frequently hear people say that things like the relational model is not suitable for capturing the structure of text, or referring to text by the peculiar term “unstructured data”. Some writers talk about combining text and relational data, as if there were a contradiction. As if relational data were a special kind of data. A more correct account, like that of Curt Monash, is to simply note that text does not work very well with the architecture of mainstream relational products – which is true.
Since text search is one of my top areas of expertise, I hope I can explain to you why the relational model is perfectly capable of capturing the structure of text. I'll start at the very bottom, explaining what text search really is.
Eventuellt bör nämnas att jag känner Jesper, han är en både smart och klok (samt trevlig) person.