Re: storing survey answers of different data types

From: paul c <toledobythesea_at_oohay.ac>
Date: Fri, 24 Apr 2009 02:41:18 GMT
Message-ID: <iF9Il.23863$Db2.16258_at_edtnps83>


Joe Thurbon wrote:
> On Fri, 24 Apr 2009 10:17:57 +1000, Bob Badour
> <bbadour_at_pei.sympatico.ca> wrote:
>

>> Joe Thurbon wrote:
>>
>>> On Fri, 24 Apr 2009 03:32:22 +1000, Bob Badour 
>>> <bbadour_at_pei.sympatico.ca>  wrote:
>>>  [...]
>>>
>>>> My name is Bob, I have property in Canada, my house is 114 years old.
>>>> {name=Bob,place=Canada,age=114)
>>>>
>>>> Different questionnaires. Different tables. A column for each 
>>>> question.  A row for each respondent. All described neatly in the 
>>>> system catalog.
>>>  My name is Joe, I have property in Australia, my house if 40 years old.
>>>  Just wondering, if one of the requirements for a system included
>>> something like 'Be able to list all questionnaires', would
>>> you still consider one-table-per-questionairre a reasonable design?
>>
>> Absolutely. It's a simple query from the system catalog.
>>

>
> Am I right in saying that there is no standard structure for the
> system catalog? (Not that that is really germane in a theory newsgroup)
> ...

I think it is germane, not to mention relevent. But something has to give when the system's devices start talking about the system, eg., I vaguely remember a system that insisted on a redirection operator for equating names in system tables with actual tables. Without that you need some kind of other language device to 'switch gears', maybe all I'm saying is that one always needs some kind of language device to 'switch gears'. I think it is better to have a 'device' that puts one in the 'mode' of the ordinary user, that way the 'device' follows user rules and doesn't stray into mysticism.

>>
>>> I think that there is a more abstract question trying to get out
>>> of my head. Maybe it's: 'When relations become things that
>>> have facts asserted about them, should one stop treating them as
>>> relations, and normalise further?" (where normalise is almost certainly
>>> the wrong word, but I'm not sure what the right one is.)
>>
>> You must not be phrasing that well. All interesting relations have 
>> facts asserted about them. Degree. Cardinality. Functional 
>> dependencies. etc.

>
> Of course, you are right.
>
> Which unfortunately means that my question is now going to have to be
> asked as a series of problems.
>
> (I've just snipped such a problem from this post, because my reply
> to Gene in this thread ended up being a more succinct description of it).
>
> Cheers,
> Joe
Received on Fri Apr 24 2009 - 04:41:18 CEST

Original text of this message