Re: Nearest Common Ancestor Report (XDb1's $1000 Challenge)

On 6 Jun 2004 14:16:38 -0700, Neo wrote:

>> In your original post (with classes only, no subclasses)
>
>The statements...
>person isa thing.
>john isa person.
>
>...create a class hierarchy that can be viewed as:
>thing
> person
> john
>
>> storing that each class is a thing would be redundant.
>
>Yes, that would be redundant, and I didn't, nor am I asking you to,
>but that fact should be derivable. Upto RM Sol#2, a db client cannot
>derive that fact as one can with XDb1's db, because the provided
>solutions don't store the class of some things.

Of course that can be derived. You are (again) showing your lack of SQL knowledge. Since you admitted that each thing is a thing and that storing that explicitly would be redundant, you'll have no trouble with this simple query (based on RM#2's table structure):

 SELECT thingName + 'isa thing.'
 FROM things

>> But since your original message had no hint towards subclasses,
>> that's a moot point anyway.
>
>"any hierarchy" includes the class hierarchy.

If there are no subclasses, there is no class hierarchy.  

>> irrelevant, since your original post had no mention of a class hierarchy.
>
>"any hierarchy" includes the class hierarchy.

See above.

And also see my messages of yesterday. I already agreed to an arbitrator and I promised that I would change RM#2 to make creation of the report for a class hierarchy possible as well IF the arbitrator thinks I should (even though XDb1 is not able to do this).

Best, Hugo

-- 

(Remove _NO_ and _SPAM_ to get my e-mail address)