Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> comp.databases.theory -> Re: The horse race

Re: The horse race

From: Mark Johnson <102334.12_at_compuserve.com>
Date: Sat, 25 Feb 2006 11:43:40 -0800
Message-ID: <sab102to715rs7dlrktb354godcls8c6ck@4ax.com>


"Tony Andrews" <andrewst_at_onetel.com> wrote:

>Mark Johnson wrote:
>> "Tony Andrews" <andrewst_at_onetel.com> wrote:

>> >Mark Johnson wrote:

>> >Sorry to butt in, but I have just been read this and another related
>> >thread for the first time, and it is getting bizarre. Mark, are you
>> >asking Marshall to attempt to count a collection of some kind (punch
>> >cards, lines of code, whatever) when they are not in their correct
>> >order? And are you implying that he won't be able to do it?

>> I'm pretty sure. Some of those punch card programs ran hundreds of
>> cards. As cliched as the joke is, I'm sure one or more people dropped
>> their boxes of cards on the way to the old central computer card
>> readers. It's a lot to resort, if, assuming if, the cards were not
>> otherwise numbered in order.

>But the key word here was COUNT. You are saying that items have to be
>"in order" to be able to COUNT them?

You just said that. That's not even barely implied in anything you quoted. The issue was whether or not data, in this case a set of punched cards, were in their proper order, and therefore useful.

>> The problem appears to be that in order to think in terms of sets and
>> the 'new math', that some are driven to defend it to an absurd degree,
>> claiming that proper order is essentially irrelevant. And I've simply
>> tried to offer numerous examples showing that proper order is
>> essential to most all data. However accounted, in a database, that
>> order must be accounted. Other examples were a roster of US
>> Presidents, start and finish positions in a horse race, literally the
>> order of words and phrases in these very messages back and forth, and
>> so on.

>I am sure nobody sensible (and I know that includes Marshall) would
>claim that order is never important.

He has. That's what was so incomprehensible. The idea was apparently to defend some notion of disorder, or at least the irrelevancy of proper order. Sets are said to be unordered, therefore order is irrelevant, something like that. I told him that's not necessarily what the 'new math' says. So it led to a defense of disorder to the extent of posting stanzas from a poem, but completely out of order, and then saying it looked just fine - no harm, no foul. And I offered a host of counter-examples, not the least of which were these very messages if you take into account the order of the paragraphs, and the order of words and phrases within each sentence. I took to rearranging some and pointing out that even if they seemed to mean something semantically, or were barely syntactically, correct that at the very least this was not the original meaning. That data was lost.

But as you say, this is something so patently obvious that it - need not be said. I was stunned to see that sort of thing, not in one message, but in entire subthreads.

>I think what is claimed is that
>(a) sets are not intrinsically ordered, and (b) sets can nevertheless
>be used to store data which may then be ordered, perhaps in more than
>one way. For example, you could list the horses in the race in order
>of finish position, or in order of horse's name - all from the same
>intrinsically unordered set of data. Do you not agree?

But that goes to my original question. If you have a binary relation of President's surname and the order in which he served, do you still insist that such is intrinsically unordered, or that it is simply an ordered set? It seems like semantic wordplay, to me. You say the table cannot be ordered because it's a set. You say a table with nothing but order and ranking is a set, and therefore is not ordered. It's conveniently self-referential. It's a Catch-22. I say, if it looks to be in order, even if one could treat it otherwise, if the data only makes sense in that particular order, then it can be said to be intrinsically ordered. And you say, no, see above.

I would agree that one could reductively, but only reductively, claim that the internal representation in storage, the machine representation, is unordered. That's only obvious. One could store P, from the Pyramids, over here, and "mids", somewhere else, for that matter. But the machine doesn't understand the semantics, and only the syntax which is necessarily programmed into it for this task. It's a practical matter. It will change with the machine that's used. The bits of information could be spread out and dissected in any order or 'bank', whatsoever. You point to the cards fallen on the floor and say - see. It's all there. But that's exactly the problem. I say that they have become useless unless placed in proper order. That order is intrinsically part of the data, and the data cannot be represented in any other order. If it is, the program won't run. At best you could use it to count cards, or for some other purpose, such as counting various frequencies among cards. But - the thing, the entity - is no longer that thing. You may say, sure it is. And there we are. Received on Sat Feb 25 2006 - 13:43:40 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US