Re: Announcing New Blog

"Christopher Browne" <cbbrowne_at_acm.org> wrote in message news:m3slrnsfd8.fsf_at_mobile.int.cbbrowne.com...
>> Database systems is an evolving environment.
>> Date's pedagogy on the RM is not.
>> It has been static for nigh on 30 years.
>
> The thing is, pedagogy needs to connect with theory.
>
> I don't see that relational theory has materially progressed in the
> last twenty years.

It hasn't.

> In a way, it's like the challenge of getting a Ph.D in Shakespearean
> literature. Doing a Ph.D requires having something new and novel to
> say about the area of study. As you approach 500 years since the
> death of the author, it gets increasingly difficult to find anything
> new.

Database systems are not the product of one author and are constantly evolving; the theory is not keeping pace.

> I don't see that anything much new has come up lately that would cause
> Date's pedagogy to need to change. Every couple years I browse TODS
> and basically see nothing that seems of practical importance, and even
> less that extends the RM.

Database systems since the late 90's have acquired a class of objects termed stored procedures. Essentially these are program objects written in the SQL of the RDBMS and are stored within the database.

There is in progress a migration of lines of code from the application software environment (external to the database) to the stored procedure environment (internal to the database) and this needs to be recognised.

We need a theory not just of the data, but of the data and its processing, which will address stored procedure objects (ie: processes) which are today, and not 30 years ago, capable of being stored and managed within the database systems environment.

Here are some relevant articles with further info:

http://www.mountainman.com.au/software/history/relational_model_incomplete.htm and
http://www.mountainman.com.au/software/Theory_of_Organizational_Intelligence.htm

-- 
Pete Brown
www.mountainman.com.au