Re: I think my book may be wrong about cardinality, but I'm not sure
From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Fri, 27 Jul 2007 00:07:21 -0300
Message-ID: <46a9613c$0$8841$9a566e8b_at_news.aliant.net>
>
> ...
>
> ...
>
> Not much to be found about Table_Dee and Table_Dum there. I believe
> Hugh Darwen coined the terms fifteen or more years ago and wrote about
> them in one or more articles that CJ Date re-published in his columns
> and books ("Writings" series). Darwen mentioned somewhere that he
> didn't "discover" these relations, that one or more people working on
> PRTV or ISBL did, IIRC, perhaps it was a guy named Stephen Todd. They
> seem important to me, so I'd be a little curious to know who first came
> up with them or whether it was one of those notions that occurs to
> bright groups of people once in a rare while. Again, my memory could
> well be faulty, but I seem to remember Darwen mentioning that he had
> brought up tables with no columns or relations with no attributes or
> something like that to Codd and Codd had dismissed them. I wonder
> whether his reasons had more to do with practice than theory?
>
> Anyway, if a set can "be" and yet not have members, it certainly seems
> that a relation with an empty header is possible. And DEE is an obvious
> identity for join, eg., A <AND> DEE = A. I believe that in TD, DEE is
> represented as "TRUE". Try and put that "tuple" in your N-dimensional
> space!
>
> Some of the columns about these tables were published in a magazine
> called Database Programming and Design, DBPD for short, which is now
> called "Intelligent Enterprise such and such" or similar oxymoron. Its
> online archives might still have some of Date's versions of Darwen's
> papers. I remember when I first saw them and lights started to go off
> in my head because they were so approachable, like a lot of those old
> columns, short and sweetly sensible with a minimum of technical lingo.
> Reading that magazine was an experience somewhere between schizophrenia,
> psychodelia and maybe Philadelphia because I'd first look at Date's
> column, accidentally flip a couple of pages and be confronted with one
> of the several other columnists' mumbo-jumbo. I always figured the
> publishers had made some mistake, kind of the reverse of Mr. Bean being
> hired as an art historian.
>
> I hope Bob B will correct anything I've got wrong above. It seems like
> a month since he's corrected anything of mine but it's been a lifelong
> (so far) trait of mine that I can't go that long without saying
> *something* wrong. Perhaps he's losing interest like Fabian P.
>
> p
Date: Fri, 27 Jul 2007 00:07:21 -0300
Message-ID: <46a9613c$0$8841$9a566e8b_at_news.aliant.net>
paul c wrote:
> Bruce C. Baker wrote:
>
>> "Bob Badour" <bbadour_at_pei.sympatico.ca> wrote in message >> news:46a90c7e$0$8837$9a566e8b_at_news.aliant.net... >> >>> Bruce C. Baker wrote:
>
> ...
>
>>>> >>>> I think I see where you're going with True and False as defined in >>>> your post, though ... >>> >>> >>> It's not so much where I am going as it is where the relational folks >>> have already gone. In particular D&D. >> >> >> >> In the Third Manifesto, 3/e? >>
> ...
>
> Not much to be found about Table_Dee and Table_Dum there. I believe
> Hugh Darwen coined the terms fifteen or more years ago and wrote about
> them in one or more articles that CJ Date re-published in his columns
> and books ("Writings" series). Darwen mentioned somewhere that he
> didn't "discover" these relations, that one or more people working on
> PRTV or ISBL did, IIRC, perhaps it was a guy named Stephen Todd. They
> seem important to me, so I'd be a little curious to know who first came
> up with them or whether it was one of those notions that occurs to
> bright groups of people once in a rare while. Again, my memory could
> well be faulty, but I seem to remember Darwen mentioning that he had
> brought up tables with no columns or relations with no attributes or
> something like that to Codd and Codd had dismissed them. I wonder
> whether his reasons had more to do with practice than theory?
>
> Anyway, if a set can "be" and yet not have members, it certainly seems
> that a relation with an empty header is possible. And DEE is an obvious
> identity for join, eg., A <AND> DEE = A. I believe that in TD, DEE is
> represented as "TRUE". Try and put that "tuple" in your N-dimensional
> space!
>
> Some of the columns about these tables were published in a magazine
> called Database Programming and Design, DBPD for short, which is now
> called "Intelligent Enterprise such and such" or similar oxymoron. Its
> online archives might still have some of Date's versions of Darwen's
> papers. I remember when I first saw them and lights started to go off
> in my head because they were so approachable, like a lot of those old
> columns, short and sweetly sensible with a minimum of technical lingo.
> Reading that magazine was an experience somewhere between schizophrenia,
> psychodelia and maybe Philadelphia because I'd first look at Date's
> column, accidentally flip a couple of pages and be confronted with one
> of the several other columnists' mumbo-jumbo. I always figured the
> publishers had made some mistake, kind of the reverse of Mr. Bean being
> hired as an art historian.
>
> I hope Bob B will correct anything I've got wrong above. It seems like
> a month since he's corrected anything of mine but it's been a lifelong
> (so far) trait of mine that I can't go that long without saying
> *something* wrong. Perhaps he's losing interest like Fabian P.
>
> p
Hey! I don't correct what you write. I just clarify it a little from time to time. Received on Fri Jul 27 2007 - 05:07:21 CEST