Re: More on lists and sets

From: David Cressey <dcressey_at_verizon.net>
Date: Fri, 24 Mar 2006 16:15:17 GMT
Message-ID: <pWUUf.2$ui7.1_at_trndny09>


"Jon Heggland" <jon.heggland_at_idi.ntnu.no> wrote in message news:e012ql$88i$1_at_orkan.itea.ntnu.no...
> David Cressey wrote:
> >> Glad to hear it. Note that no lists or "NFNF" data is involved. :)
> >
> > I'm confused. How does dataphor represent and manipulate tree
structured
> > data?
>
> With relations. I'm not sure I understand the question. Of course, the
> relations are presented graphically using an table widget where the
> tuples must necessarily be presented in some order---so you could say
> there is a list involved at that level..?

Here's what I mean:

In SQL, tree structured data can be represented by adjacency lists, or by nested sets. Both of these, to some extent, can be seen as a "kludge" by those who are accustomed to dealing with tree structured data as it is represented in lisp, or in the Pick class of products as presented by Pick enthusiasts in here.

Calling it a kludge doesn't mean that it doesn't work. It does work. It means that it solves a problem that wouldn't exist in the first place in some other environment.

 I hasten to add that I've derived benefit from both the adjacency list and the nested set techniques. They are in my toolkit. And there are things I do without "stepping outside of SQL" that other people might imagine it's impossible, or nearly so, within SQL.

But just because the RDM (and its bastard offspring, SQL) can represent all data, in any context, doesn't mean that it's the appropriate tool in every context. I don't want to be a kid with a hammer called RDM (or SQL) any more than I want another person to be a kid with a hammer called UML.

As I've watched the battle play out between "relational purists" and people like Dawn in this newsgroup over the last couple of years, there seems to be a common subtext in both of their positions. It can be stated roughly as: "there isn't room enough in this town for both of us." My position is that, yes there is room enough. Received on Fri Mar 24 2006 - 17:15:17 CET

Original text of this message