Re: cdt glossary - TABLE

From: mAsterdam <mAsterdam_at_vrijdag.org>
Date: Mon, 11 Jul 2005 02:09:38 +0200
Message-ID: <42d1b8c3$0$10317$e4fe514c@news.xs4all.nl>

Paul wrote:

> mAsterdam wrote:
> 

>>Lists are ordered by definition
>>(http://en.wikipedia.org/wiki/List), so
>>
>> "the table header is an (unordered) set of column names"
>>
>>directly contradicts
>>
>> "the table header consists of a list of column names"
>>
>>I don't have a problem listing contradictory definitions
>>if they are in use somewhere (in database context).
>>However, this is the first time I see a table header
>>with *un*ordered column names mentioned.
>>Are you sure? Could you provide some background?

> 
> I thought the whole idea of the relational model was to abstract away
> from physical things like the order of the columns, which isn't relevant
> to data management.

I think so to - but we are talking about confusions around the word 'table' here. It is used in daily speech and in SQL, but is it (still) part of the relational model?

> I think it's done this way in one of Date's books, maybe the Third > Manifesto?

A quick lookup gave me these quotes:

C. Date about "SQL Facilities" in Intro, 8th ed, p 161:

	"Within a given row, the component values ... are
         thus identified primarily by their ordinal position
	(even when they also have names,
	which is not always the case)."

and (with H Darwen) in TTM, 1st ed, p135, footnote:
	"/Relation/ has a precise (and somewhat abstract)
         definition; /table/ by contrast, does not."

  (italics original)

The first quote indicates that "the table header is an (unordered) set of column names" is _not_ how SQL looks at 'table'. I didn't check thoroughly though, so I may have missed contrary remarks. If so, please provide them. Received on Sun Jul 10 2005 - 19:09:38 CDT