That is a good point! The right plan in the compound index could have made
all the difference.

Ivan Prenosil wrote:

>>From: Thomas Miller
>>The guy created a third column and added column 1 and column 2.
>>
>>By using the third column, he found a considerable performance boast
>>versus using indexes on Column 1 and 2, or a combination index
>>that included both.
>>
>>
>
>Yes, this was already written in the message from b.p.i.sql, but ...
>what conclusion could you make of it ? That replacing compound index
>by index on single concatenated column will make your queries 500 times faster ?
>I am not saying it is/is not true/possible, just that there are not enough informations.
>How can you tell that the speed difference was caused by "more efficient indexes",
>and not just different plans ?
>I would not make any judgements without more testing, comparing plans/statistics etc.
>
>Ivan
>http://www.volny.cz/iprenosil/interbase
>
>
>
>To unsubscribe from this group, send an email to:
>ib-support-unsubscribe@egroups.com
>
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
>
>

--
Thomas Miller
Delphi Client/Server Certified Developer
BSS Accounting & Distribution Software
BSS Enterprise Accounting FrameWork

http://www.bss-software.com



[Non-text portions of this message have been removed]