Re: Multi Valued Interface Models?

Marshall Spight wrote:
> dawn wrote:
> > ... separate yourself from thinking about the data model for the
> > database (that is the model of the interface to the database) and focus
> > on the data model of the interface to the user. It simply is not
> > possible for the RM to be the data model for that interface.
>
> Can you say why? I read your blog posts, and I didn't see any
> justification
> for this statement. Your example UI model wasn't very enlightening,
> either.

I thought I said why. If you take an implementation of the RM into a language, such as SQL (or some perfect version thereof), you cannot have a view of data in that language that is the data interface of this UI. While you can unnormalize data for SQL views in other ways, you cannot get them out of 1NF. An arbitrary user interface is not a 1NF view. This does not mean that your data cannot be stored in a database where the interface to the database is RM, but that the data model of the UI cannot be the RM.

> In any event, I am certain that the "not possible" part of the above
> statement
> won't hold up, because the RM can model any concrete type, even if
> restricted to 1NF.

How would you model the user interface (not some database related to it) in the example given using a language that implements the RM?

> I'm not saying "convenient" necessarily, just
> possible.

Are you talking about adding list types to the RM?

> On the other hand, you use the words "model" and "interface" in a
> lot of different ways, and it's possible I'm not following your
> argument.

Model is a very broad term, but I tried to use "data model" consistently in line with the definitions a couple of blogs earlier.

> Calling out specific concerns is always helpful to the conversation.
> I sometimes feel like you spend too much time on definitions,
> especially
> definition of multivalued (ha ha) words like "interface". Specific
> terms
> are easier to define than generic terms.

So you are saying that it is OK that I didn't define "interface" or that I should have defined it (or that I did define it?) What I found when reading about the RM was that whenever I saw a gap in reasoning, the definition of the RM jumped out of the way of it. For some of this discussion it is more important that a def stay fixed long enough to look more closely at it than precisely what that definition is.

I clearly still haven't articulated this really precisely. Hmmmm. --dawn Received on Thu Feb 09 2006 - 23:05:46 CET