[AccessD] Number vs text data type

If you do something a million times and measure the total time/iterations 
the accuracy improves.  But I don't see the relevance anyway.  Indexes are 
numeric pointers into the data anyway so the time to find a number or string 
that it indexed involves only a few comparisions.  The big difference is in 
the size of data and index that needs to be transferred from disk to 
processor which takes far more time than the time to make the comparisons.  
Once the table or index is cached, the processor speed makes very little 
difference.  The real issue is the fact that the data is much larger and the 
index is also as much larger and with a file server database, the size of 
the data and index makes far more impact on the ultimate performance than 
the processor speed.  I've recently seen a 200 mhz and a 2.4 ghz computer 
load the same data from a file server on the same LAN in the same time.  Yes 
the faster machine filters and sorts much faster.  You should see that 200 
mhz machine with cable modem kick the butt of the 2.4 ghz machine using a 
cellular phone modem loading web pages.  Multiplying the processor speed by 
12 times makes no discernable difference in load time, but there is no doubt 
that cutting the size of the data in half will improve data load time by a 
factor of two no matter what the speed of the processor.  You can only buy 
so much performance.  The rest has to be consciously designed.  Also, things 
like spaces, parentheses and hyphens can be used in formatting a numeric 
value.  Why do you suppose Microsoft and other software vendors store dates 
as numeric values?  Speed, efficiency and format flexibility.

Hen



>From: "Stuart McLachlan" <stuart at lexacorp.com.pg>
>Reply-To: Access Developers discussion and problem 
>solving<accessd at databaseadvisors.com>
>To: Access Developers discussion and problem 
>solving<accessd at databaseadvisors.com>
>Subject: RE: [AccessD] Number vs text data type
>Date: Thu, 16 Oct 2003 08:16:21 +1000
>
>On 15 Oct 2003 at 14:46, Heenan, Lambert wrote:
> >             NumericSearchStart = Timer
> >             .FindFirst sCriteria
> >             intHits = intHits + Abs(CLng(Not .NoMatch))
> >             NumericSearchEnd = Timer
>...
> >             TextSearchStart = Timer
> >             .FindFirst sCriteria
> >             strHits = strHits + Abs(CLng(Not .NoMatch))
> >             TextSearchEnd = Timer
>....
> >
> > And here are the results of a few runs, last to first...
> >
>Average find times
>Numbers:0.0020859370 Hits:0
>Strings:0.0025468750 Hits:0
>
>The results are meaningless :-(
>Timer is only accurate to about 1/18th of a second ie to .0556
>seconds. Any test which takes less than that will either return 0 or
>.0556.
>
>You are carrying out a single indexed lookup on an open table plus
>a calculation using 5 functions (addition, Abs() CLng(), negation,
>NoMatch)
>
>Which do *you*  think takes longer (ie has most effect on your
>results)?
>
>
>--
>Lexacorp Ltd
>http://www.lexacorp.com.pg
>Information Technology Consultancy, Software Development,System
>Support.

_________________________________________________________________
Add photos to your e-mail with MSN 8. Get 2 months FREE*.  
http://join.msn.com/?page=features/featuredemail