Re: Announcing New Blog

David Cressey wrote:
> "Jay Dee" <ais01479_at_aeneas.net> wrote
>
<snip>
> What Dawn is doing is to try to gather the available data to try to explain
> what she sees as a disconnect between theory and practice in the comparison
> between database services centered around the RDM and parallel services
> centered around alternative formulations.
>
> I'm very critical of Dawn's writings, but that's because I don't perceive
> the same disconnect as she does, and I regard my own background as
> extensive enough so that I don't feel compelled to take her word for it that
> there is such a disconnect. In other words, Dawn and I disagree at the
> level of raw data, or at least personal experience.
>
> Dawn is proceeding on the assumption that the raw data of which her own
> experience is illustrative

A minor correction on that. I proceeded with my investigation into data modeling after working with SQL and non-SQL-based tools and recognizing a pattern of the SQL implementations being more expensive in real-dollars-in-my-budget terms. I did not assume that my experiences were the same as others. So, that is what prompted my research, but I recognize it is purely anecdotal as evidence.

> will eventually lead to a theory of software
> development that will explain what she perceives as a demonstrated fact,

This is not correct. I have mentioned several times in this forum that I do not know how to set up an experiement and there is no emperical data from which I can derive a conclusion on whether SQL-based or even RM-based solutions are a better bang for the buck than some environments that did not arise from the RM. I do not claim and see no path at this point to claiming that my opinion is a demonstrated fact.

Now that I have given what you accurately referred to as an opening statement, I will proceed to introduce both theory and practical evidence, non of which I consider conclusive, before giving a closing argument, likely in Q4 this year. It will be up to the reader to decide whether the RM is the only real database theory, as many of its proponents claim, whether or where it might be a best practice, and whether there is reason for the industry to move away from seeing the RM as king of the hill in data modeling (and college texts).

> one
> that I don't accept as given.

I, too, do not accept it as a given. I also don't have an airtight case for walking completely away from the RM, but hopefully more compelling than the case for using the RM as (almost) the exclusive means for getting from conceptual to logical data models.

Cheers! --dawn Received on Tue Jan 17 2006 - 17:38:54 CET