Re: Date, Darwen, Pascal and the alternative to Nulls in the RM

"Tony Andrews" <andrewst_at_onetel.com> wrote in message news:1143023929.620776.83400_at_i40g2000cwc.googlegroups.com...
> mountain man wrote:
>> The only way to avoid nulls is to avoid change management.
>> That's it really. End of story. Change management will eventually
>> breed the occurrence of nulls.
>>
>> As change management is not incorporated into their theories,
>> they are able to sustain a static rant endlessly. Nothing much has
>> changed in the 30 years concerning their RM mantra.
>
> Actually, Fabian Pascal covers this very point in his book Practical
> Issues In Database Management. IIRC his solution involves creating a
> new table each time the schema changes. I'm not saying you'd like his
> ideas, but they do exist.

Well that's an encouraging sign, really.

>> The above theorists are too busy elsewhere chasing the
>> elusive non existent nulls, and writing commentaries
>> upon commentaries upon commentaries.
>>
>> What's new?
>
> I agree that theorists work in a pure, abstract domain not sullied by
> the difficulties faced in the "real world" by people who have to "get
> their hands dirty". But surely you don't think we don't need
> theorists? If we all just made the best of today's products and didn't
> try to envisage the ideal solution, we wouldn't have the relational
> model, or even SQL with all its flaws, at all!

Theorists are required who understand that the RM is a stepping stone to a more complete theory of data and code, of change management and of the organisation itself.

I object to static pedagogy as a theory. A proper theory needs to grow with the environment of its scope, and as times change the practices of management of DBMS's.

Thanks for pointing out Pascal's attention to the change management issue. Although it is only one of several steps that need to be taken, it is encouraging to see this step addressed systematically.

-- 
Pete Brown
www.mountainman.com.au