Ivan, aggregative tests

[Ruslan Zasukhin]

on 12/11/03 11:21 AM, Ruslan Zasukhin at sunshine at public.kherson.ua wrote:

> 2) By standard 
> 
>       SELECT f1, f2, f3, f4, min(f)
> 
>   is wrong. Standard say we cannot mix aggregative and normal columns.
>   
> But I believe this is not very correct.
> IF we have only min/max in the SELECT, then we do not have ambiguity, and we
> still can correctly show other fields of found min/max record.
> 
> Okay, here must be even more strict!
> 
>       SELECT f1, f2, f3, f4, min(f), max(f)
> 
> This already is wrong, because min an max will choose DIFFERENT records.
> And we cannot show f1, f2, f3.
> 
> So we can DIFFER from standard, IF select have only min() or only max()
> function. 
> 
> Agree?

I have to think more here...

1) this is also wrong

       SELECT f1, f2, f3, f4, min(f3), min(f4)
                              ^^^^^^^^^^^^^^^^^ different records.


So correct can be SELECT with ONLY one aggr column of type min or max.


2) but I have try to implement this check, and it is very hard.
Too big overhead for analyze.

So, Ivan, let's drop this idea for now.


-- 
Best regards,
Ruslan Zasukhin      [ I feel the need...the need for speed ]
-------------------------------------------------------------
e-mail: ruslan at paradigmasoft.com
web: http://www.paradigmasoft.com

To subscribe to the Valentina mail list go to:
http://lists.macserve.net/mailman/listinfo/valentina
-------------------------------------------------------------