Re: foundations of relational theory?

From: Mike Preece <michael_at_preece.net>
Date: 2 Nov 2003 20:16:11 -0800
Message-ID: <1b0b566c.0311022016.80f88d5_at_posting.google.com>


byrmol <member27348_at_dbforums.com> wrote in message news:<3551203.1067812771_at_dbforums.com>...
> > Basically:
> > * Structure:
> > All data in D3 is stored as items within files.
> > Items within these files are divided into sets called attributes which
> > contain one, multiple, or no values.
> > * Integrity
> > Can be enforced in multitudinous ways. It is perhaps worth noting that
> > much of what is elsewhere called "data integrity" falls (quite rightly
> > imo) under the heading of "input validation" in Pick applications that
> > run wholly or partly within the DBMS. Pick has full ACID compliance
> > for transaction handling btw.
> > * Manipulation
> > What is there to say? You read, write & delete data as you need to -
> > and you need to do much less of it in Pick than with any of the
> > attempts to implement the relational model.
> >
> > If you want, as I suspect you might, more in-depth answers then can
> > you point me to a definition of a model with these sub-headings that I
> > can use as a kind of template?
> Here is the relational models conformance to a definition of a
> data model.
> Its general theory is set theory and predicate logic expressed as...
> * Structure - R-tables
> * Integrity - Domain, attribute, relation, database
> * Manipulation - R-operations (R-algebra & R-calculus)
> From your reply I am getting the impression that MV/PICK is not a data
> model at all but simply an application.

If I can have another go at this - although it still won't satisfy your requirements for a definition - it might help nonetheless...

All data in D3 is stored as items within files. Items within these files are divided into sets called attributes which contain one, multiple, multiple subsets, or no values.

System file - a number of groups of frames containing items, some of which have attributes of base address and the number of groups of frames in each==>
Master Dictionary - a number of groups of frames containing items, some of which have attributes of base address and the number of groups of frames in each==>
File Dictionary - a number of groups of frames containing items, some of which have attributes base address and the number of groups of frames in each==>
File Data-level - a number of groups of frames containing zero or more==>
Items - variable length strings, identified by a unique item-id, comprised of==>
Attributes - variable length strings, delimited by "attribute marks", which can be comprised of zero or more==> Multivalues - variable length strings, delimited by "multivalue marks", which can be comprised of zero or more ==> SubValues - variable length strings, delimited by "sub-value marks".

Items can be accessed in various ways, including: ¨ Direct hashed access, using the item-id and the base address and number of groups in the file to identify the particular group, and then via sequential search within the linked-frames that make up the group;

¨ Sequentially in the hashing sequence;
¨ Sequentially in a sorted sequence;
¨ By index key.

Various proprietary and other methods exist to allow access: ¨ To pseudo-items by "data abstraction layer" subroutine calls.

Phew! Someone else want to take over??? I gotta get some work done! Received on Mon Nov 03 2003 - 05:16:11 CET

Original text of this message