Re: Constraints and Functional Dependencies

paul c wrote:
> mAsterdam wrote:
>> Mp (Marshall/paul) referential integrity =def=
...
>> (using レ = katakana re for references)
>>
>> R(a) レ S(b) ≡
>> ∀R(a): ∃S(b)| a=b
>>
>> Classic referential integrity (no need, says paul c)
>>
>> (ル = ru for references unique)
>>

(ii)
>> R(a) ル S(b) ≡
>> ∀R(a): ∃!S(b)| a=b
>>

> ... Congrats.

Thanks.

Hmm...

Note (afterthought):
(ii) doesn't need that b should be /defined/ as ck for relvar S (which would really be the classic ri), just that it /could/ be ck in the current relvalue S.

About the notation:
Note (not afterthought):
1. that the katakana re, レ for _reference_ could read as 'check' (it looks like a check mark),
verbose 'check that there is',
and
2. the katakana ru, ル for _reference_unique_ as 'check one' (it looks like a check mark with a sloppy 1 before it), verbose 'check that there is only one'.

How about that for sugar coating ;-) Received on Tue Feb 27 2007 - 18:07:35 CST