RE:Optimisation of SELECT

>> trish mcelvaney <trish@paragon.ie> writes >> >> I am writing an application for a client (Informix 4gl) >> which will require a very large SELECT using a number >> table joins (up to 12 - each table having a large volume >> of data, approx 200,000 rows each ) and a wide variety >> of selection criteria. >> >> Srtuff Deleted... >> >> If I go for the "one select" option, is the access time >> from the database faster depending on the order of the >> tables in the FROM clause ? >> >> Any suggestions ? >> >> Trish Hello Trish, It is tempting chaos to consider doing multiple joins with large tables and a wide variety of selection criteria. I have nightmares where the database looks like a very small animal searching through a haystack looking for needle to compare to another needle in another haystack which hasn't been searched yet. I always wake up before the first needle is found. I often wonder at the genius of some programmers who can elegantly place complex logic into seemingly simple steps and I would like to think that the ONE SELECT was achievable. That said, GO FOR THE TEMP TABLES SOLUTION !! Best of Luck, Tom Hogarty < STANDARD DISCLAIMERS APPLY > Siemens Nixdorf Informationsystems Paderborn, Germany Tel : +49-5251-8-11602 ($WORK) Tel : +49-5254-65553 ($HOME) e-mail : hogarty.pad@sni.de