Re: Multi Valued Interface Models?

From: dawn <dawnwolthuis_at_gmail.com>
Date: 10 Feb 2006 17:49:27 -0800
Message-ID: <1139622567.610740.164410_at_o13g2000cwo.googlegroups.com>


Jan Hidders wrote:
> dawn wrote:
> >
> > I'm working on how to state the argument another way. If anyone has
> > suggestions, please jump in.
>
> How about: "The flat relational model can sometimes be a relatively
> lousy data model when it comes to modelling the data at the external
> level, i.e., the user interface, since there we often see the data
> represented in a nested and ordered fashion."

My claim, however, is that given that any software interface can be modeled using a "data model" as implemented in a programming language, it is not possible to use the RM as the sole data model for the software interface to a user. There is no collection of statements that can be executed in a language that implements the RM (solely) that can yield the data view for an arbitary UI.

I'm not getting this through however, so either a) My definitions and logical steps are not tight enough to make this clear to most readers (I suspect that is the case) or b) I'm wrong (always possible, but I think I have a real point on this one that I'm just not getting through adequately). Cheers! --dawn Received on Sat Feb 11 2006 - 02:49:27 CET

Original text of this message