Re: Proposal: 6NF
From: paul c <toledobythesea_at_oohay.ac>
Date: Sat, 30 Sep 2006 02:26:29 GMT
Message-ID: <pBkTg.67714$5R2.2051_at_pd7urf3no>
>> paul c wrote:
>>
>>> Karen Hill wrote:
>>>
>>>> J M Davitt wrote:
>>>>
>>>>> Karen Hill wrote:
>>>>>
>>>>>> 6NF would be a database that uses no Nulls.
>>>>>
>>>>> Um, "no nulls" is necessary for 1NF. And I
>>>>> believe someone already has dibs on "6NF."
>>>>>
>>>>
>>>> How so? Plenty of people have nulls in 3NF. How is no nulls necessary
>>>> for 1NF?
>>>>
>>>
>>> I think JM is saying nulls aren't possible in 1NF, ergo not either in
>>> 3NF. Just because somebody thinks they have 3NF with nulls, doesn't
>>> mean they do, no matter what their SQL product documentation tells them.
>>
>> I guess it would have been a little more accurate to say there are
>> some people who discredit nulls who also aren't sure if 1NF really
>> means anything at all. I'm sort of in that camp, although I sometimes
>> wonder if there really is a 1NF, could it be that part of its
>> definition be that no rva could constitute, by itself, a key. That is
>> part of my interpretation of the information principle, but I'm sure
>> some other people would find my interpretation objectionable.
Date: Sat, 30 Sep 2006 02:26:29 GMT
Message-ID: <pBkTg.67714$5R2.2051_at_pd7urf3no>
Bob Badour wrote:
> paul c wrote: >
>> paul c wrote:
>>
>>> Karen Hill wrote:
>>>
>>>> J M Davitt wrote:
>>>>
>>>>> Karen Hill wrote:
>>>>>
>>>>>> 6NF would be a database that uses no Nulls.
>>>>>
>>>>> Um, "no nulls" is necessary for 1NF. And I
>>>>> believe someone already has dibs on "6NF."
>>>>>
>>>>
>>>> How so? Plenty of people have nulls in 3NF. How is no nulls necessary
>>>> for 1NF?
>>>>
>>>
>>> I think JM is saying nulls aren't possible in 1NF, ergo not either in
>>> 3NF. Just because somebody thinks they have 3NF with nulls, doesn't
>>> mean they do, no matter what their SQL product documentation tells them.
>>
>> I guess it would have been a little more accurate to say there are
>> some people who discredit nulls who also aren't sure if 1NF really
>> means anything at all. I'm sort of in that camp, although I sometimes
>> wonder if there really is a 1NF, could it be that part of its
>> definition be that no rva could constitute, by itself, a key. That is
>> part of my interpretation of the information principle, but I'm sure
>> some other people would find my interpretation objectionable.
> > Why on earth would anyone want to proscribe the use of RVA's in keys? A > value is a value. Period.
I'll try to phrase an answer that is as terse as that question.
p Received on Sat Sep 30 2006 - 04:26:29 CEST