On Sat, Mar 20, 2010 at 6:39 AM, Charlie Coleman <[email protected]> wrote:
> At 04:26 PM 3/19/2010 -0500, Stephen Russell wrote:
>>On Fri, Mar 19, 2010 at 4:27 PM, Charlie Coleman <[email protected]> wrote:
>>
>> > Anyway, if what you're doing is a sort of once-in-a-while type thing the
>> > performance over SQL probably won't be noticeable. But if it's in a tight
> ...
>>Why wouldn't you then use the power of SQL to get you the max by
>>client for a large volume instead of jumping around ?
> ...
>
> Don't worry Steve, I wouldn't expect you to understand, seeing how you're
> an MS-koolaid drinker and all. But basically, if you know the underlying
> technology, you can find ways to optimize for certain situations. I know
> that actually understanding the technology under the hood is not the "MS
> way of thinking" but it certainly can save the day sometimes (and often
> saves a lot of time figuring out problems). I've done enough testing and
> deployed enough applications to know SQL syntax is not all-powerful
> (regardless of the backend server by the way).
-----------------

You can dance around all you want singing your little song but you
never addressed why you woudl do this on a one by one basis and not
via group by in SQL?

This is purely raw data in volume.  I fail to see how rerunning your
findmax() is in anyway superior.


> I bet it's driving you INSANE that the world is still on TCP/IP.

No Charlie it is not.


> At the end of the day, no matter what project you're working on, it's
> always good to attack a particular challenge from multiple angles. You'll
> end up with the best result in the long run. I was just providing a
> potentially different angle. Again, I know it's not the MS-way, but it is
> the responsible thing to do. They used to teach this concept in software
> engineering colleges - did you miss that class?
--------------------

Let me get this story straight.

<index-disk operations>
Find the wrong client
move pointer -1
run if block to identify if this is the correct client
</index-disk operations>

do your process

loop

Sorry but this is poor design outside of localized data.  Will not
scale to alternate backends.  Will not work for non int keys.  So
where was your engineering class, A.C.M.E. U?


-- 
Stephen Russell

Sr. Production Systems Programmer
CIMSgts

901.246-0159 cell

_______________________________________________
Post Messages to: [email protected]
Subscription Maintenance: http://leafe.com/mailman/listinfo/profox
OT-free version of this list: http://leafe.com/mailman/listinfo/profoxtech
Searchable Archive: http://leafe.com/archives/search/profox
This message: 
http://leafe.com/archives/byMID/profox/[email protected]
** All postings, unless explicitly stated otherwise, are the opinions of the 
author, and do not constitute legal or medical advice. This statement is added 
to the messages for those lawyers who are too stupid to see the obvious.

Reply via email to