Re: RM and abstract syntax trees

On Nov 13, 11:42 am, paul c <toledobythe..._at_ooyah.ac> wrote:

> I think the very idea of comparing pointers to relational constructs is
> unfortunate and mis-represents relational ideas, no matter whether one
> thinks of pointers in the more general way Bob B does or the
> machine-traditional way David C and I do. When we have join and project
> at the logical level, talk about "de-referencing" is irrelevant to a
> data design.

It is irrelevant in the general case - ie until you come across an example where it is relevant. Those messy and complex integrity constraints (that I described in another post) with the RM representation of an AST can be understood easily and economically when one associates node identifiers with pointers (whereupon those integrity constraints become implicit). That is why I say that the RM is trying hard to emulate pointer semantics.

In the case of ASTs, if someone perseveres with RM and talks about "join and project at the logical level", and ignores the pointer analogy I very must doubt they will find the design easier or more concise. I would say they have fooled themselves into thinking they are working at some higher logical level than a "low level" representation using C/C++ structs.

[snip] Received on Tue Nov 13 2007 - 06:12:45 CET