Re: Demo: Things in Hierarchies (w/o RM/SQL)

On 6 Nov 2004 10:26:37 -0800, Neo wrote:

>> Exactly what part of "The discussion whether you will pay me or not is
>> closed as far as I'm concerned" do you fail to understand, Neo?
>
>The part where your solutions were nearly twice as slow as XDb1's even
>when XDb1 is executed on 5.6X slower hardware. See measurments below.

Hi Neo,

Liar!

My final (second) submission took 11.0 ms on my computer; the version if XDb1 that was current when you set the challenge took 16 ms on that same computer for the same report. The challenge was to have the report generation no more than 2x slower; my submission was actually quicker. See message mts4b09m5olm07v0b9ppienrdk8au4md6c_at_4ax.com. Or, if your usenet provider's retention is less than 5 months, check it out on google groups: http://groups.google.nl/groups?q=g:thl2529170390d&dq=&hl=nl&lr=&selm=mts4b09m5olm07v0b9ppienrdk8au4md6c%404ax.com&rnum=101

Yes - I am aware that you went on to make new versions of XDb1, improving the speed of report generation. Good for you - but not relevant for the challenge. Microsoft is also busy building a newer version of SQL Server that will probably be quicker. Information on MS' website shows that the next version of SQL Server will actually support recursive queries (see http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnsql90/html/sql_ovyukondev.asp, under the heading "Language Enhancements"). It'll be interesting to see how this will improve querying hierarchies in SQL Server - but again irrelevant to the challenge, as SQL Server 2005 was not yet available when you set the challenge.

>In addition to being slower, your solutions were not as generic or
>normalized.

My solution was as generic as your original challenge required. It failed the requirements you added later - maybe this is a good moment to remind you of the spectacular bugs I found in XDb1 when I tried the addition requirements in your program?

My solution was normalized according to the relation model's normalisation rules. That was what you requested ("using the relational model" ... "from normalized and NULL-less data").

To reiterate - I am no longer attempting to get my prize; I will not go in an extended exchange over this again. I can only hope that your pathetic display of unfairness and total lack of sportsmanship will serve as an ample warning for any prospective customers.

And I WILL rebutt you if you start to tell plain out lies, like in the above quote about my solution being slower than yours - it is not and it has never been slower.

Best, Hugo

-- 

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