Re: NULLs: theoretical problems?

"V.J. Kumar" <vjkmail_at_gmail.com> wrote in message news:Xns999098CF461F9vdghher_at_194.177.96.26...
> "Brian Selzer" <brian_at_selzer-software.com> wrote in
> news:OMCxi.33891$2v1.16469_at_newssvr14.news.prodigy.net:
>
>>
>> "V.J. Kumar" <vjkmail_at_gmail.com> wrote in message
>
>>> The full 92/2003 standard compliance does require that deferrable
>>> constraints should be implemented. The entry/intermediate compliance
>>> level does not. There is no notion of "Core SQL" in the standard.
>>>
>>
>> FYI: I pulled "feature outside Core SQL" from the heading of Table 32
>> in ISO/IEC 9075-2:1999 and Table 35 in ISO/IEC 9075-2:2003.
>>
>
> Yes, you are quite right, I was under impression that they had preseved
> the entry/intermediate/full level taxonomy in '2003 but instead they use
> 'Core' and 'outside Core' terminology. Sorry about that.
>
> It is odd though that they consider features that are indispensable for
> any modern SQL based imlementation as not 'core'. For example,
>
> dynamic SQL
> alter table drop column/alter table add constraint/drop constraint
> isolation level other than serializable (this one is especially funny !)
> derived table ( select * from (select abc from t1) x)
>
> etc...
>

Very odd indeed. Received on Sat Aug 18 2007 - 21:23:49 CEST