Re: Modeling Data for XML instead of SQL-DBMS

JOG wrote:
> dawn wrote:
> > mAsterdam wrote:
> > > dawn wrote:
> > > ...
<snip>
> Already into a quagmire of definitions then. Look dawn, I'm personally
> not comfortable with the definition of 'data' in the term 'database',
> much to Bob's dismay at those who share my 'new-fangled' view, but what
> I did learn from him is that when we communicate it is just obstinate
> not to share established derivations. So now I happily use data to mean
> a statement of fact. Its hardly going to kill me, and it means people
> can understand each other.

I agree completely.

> Equally we all use a (the!) common
> understanding of Logical Data Model here.

I clearly missed the memo on that def and am very happy to use it. I read through and, for the most part, adopted Pascal's termiology from his Unmuddling Modeling paper due to the differences in what I had read, desiring to have a "standard" definition. I later wrote up my own understanding in a blog entry so that if there was confusion on the terms, I could be corrected.

So, what is the name of the data model specified to a DBMS tool (whether that tool uses a relational model or not)?

> There's just no point in
> being polemic - just use the consensus definition and we can all
> happily comprehend each other without 30+ posts on definitions. There
> is simply no need for renaming.

I COMPLETELY AGREE and very much want to use common definitions.

>
> Now to address your original post again, with something that may
> perhaps be useful - it occurred to me that given XML forces everything,
> naturally or otherwise, into a hierarchy, it might be worth you
> investigating something like the ACM portal for data structuring
> standards in hierarchical databases.

I have scanned through a few hierarchical database papers (before I inadvertantly ended up trading my ACM portal login for an IEEE portal login -- airheaded move, punishing myself by not paying the additional right now to correct, so next year...).

I have not landed on any "best practices" type of information, although I don't doubt that someone has put something together. I'm planning to write a blog entry on it, so I'm hoping to find something. The more recent works I have seen (typically related to XML, RDF, semantic web, etc) are more document-oriented.

> Obviously the references will be
> old because of hierarchical databases being superceded by the RM, but
> you might find something of value for you. J.

Yes, there are both old and new, but everything I'm finding on modeling data is related to the RM. Those related to designing hierarchical or codasyl databases tend not to be relevant to app software developers (the tools I'm thinking about are more like di-graphs than trees).

I have asked a corresponding question on cdp in order to collect information from experienced developers. I might ask on a cache list too. But to date I have found no developers who use more than design using normalization related to functional dependencies, intuition, experience, and proprietary standards they have established based on these.

Thanks. --dawn Received on Thu Oct 26 2006 - 14:10:20 CEST