Re: RM and abstract syntax trees

From: Marshall <marshall.spight_at_gmail.com>
Date: Wed, 31 Oct 2007 14:42:10 -0000
Message-ID: <1193841730.322245.259650_at_v23g2000prn.googlegroups.com>


On Oct 30, 5:59 pm, Bob Badour <bbad..._at_pei.sympatico.ca> wrote:
> Marshall wrote:
> > On Oct 30, 10:39 am, Bob Badour <bbad..._at_pei.sympatico.ca> wrote:
>
> >>I would like to see more heavy thinkers thinking about 6NF.
>
> > You've hinted at this idea before. I make no claims about
> > being a heavy thinker, but I've generally found your ideas
> > to be worth pursuing. Do you have any suggestions for
> > what I should be reading, before I start doing any
> > thinking? :-)
>
> I don't know about reading but a good skim of /Temporal Data and the
> Relational Model/ by Date, Darwen and Lorentzos might get some juices
> flowing. And if I recall correctly Fabian may have had a word or two to
> say in /Practical Issues.../ Frankly, I don't know whether anyone else
> has even looked at 6NF.
>
> 6NF has apparent utility for temporal data,

I gotta read that "Temporal Data" book, dagnabbit. I despair that I am such a slow reader.

> and I think it seems to have
> utility for partially known data.

Ha! We've had that conversation before. And in fact after an exuberant conversation or two (ahem) you've convinced me that I need to look at that more closely.

> At the same time, current syntaxes
> seem a little awkward when working with 6NF. The question is: What
> short-hands would facilitate working with 6NF data? What is the
> significance, if any, of those short-hands?

I am reminded of David's "curious notation" thread (was it?) from a few months ago. But I'm deferring thinking too hard about that for now.

I have this idea of developing a suite of queries to use to test out various syntaxes. In addition to SQL-inspired queries, I'm interested in culling FOL formulas from logic books. Although I notice that, for example, the notation of logic is optimized for the case that logicians have, which is expressing formulas for no particular domain, whereas we developers have an actual domain, with a great deal of specificity. I have yet to see a predicate in a logic book called "Customers" or "Employees"! :-)

Marshall Received on Wed Oct 31 2007 - 15:42:10 CET

Original text of this message