-
Re: two simple INNER JOIN questions
"Lennart" wrote in message
news:f6fa8q$6jj$1@registered.motzarella.org...
> Lennart wrote:
> [...]
> Since natural join is distributive [ (a join b) join c = a join ( b join
>
> Should have been: Since natural join is associative ...
>
But, as Lee pointed out there is no particular relationship between an
autoincrement column in one table and an autoincrement coulmn in a different
table. So, at the semantic level, there is nothing "natural" about the
joins that the OP presented as a case study.
Saying that a join across two unrelated columns both with name "id" is
"natural" is really stretching a point, in my book.
> /L
>
> [...]
-
Re: two simple INNER JOIN questions
David Cressey wrote:
> "Lennart" wrote in message
> news:f6fa8q$6jj$1@registered.motzarella.org...
>> Lennart wrote:
>> [...]
>> Since natural join is distributive [ (a join b) join c = a join ( b join
>>
>> Should have been: Since natural join is associative ...
>>
>
> But, as Lee pointed out there is no particular relationship between an
> autoincrement column in one table and an autoincrement coulmn in a different
> table. So, at the semantic level, there is nothing "natural" about the
> joins that the OP presented as a case study.
Yes, you are right. But it does not make sence to do any kind of join
between two tables with autoincrement columns, and natural is less
typing than other joins.
The OP's query was about the cardinality of the resulting table and
whether an inner join or a natural join is used is irrelevant.
>
> Saying that a join across two unrelated columns both with name "id" is
> "natural" is really stretching a point, in my book.
You seem to put some value in the word natural? AFAIK it is only a
syntactical construction in sql.
/Lennart