Gustav Brock
gustav at cactus.dk
Wed Jul 28 10:57:27 CDT 2004
Hi Mark and Scott Maybe I'm dense but - given your example data - this gives you exactly what you are asking for: SELECT bo, phone FROM tblPhone GROUP BY bo, phone HAVING Count(pro_cm_seq)=1; /gustav > The table in question uses the following fields: > bo_id...unique number string for each customer > profile_cm_seq...unique number string for each phone number > primary_ind...yes/no filed indicating primary > Effdt...date row is active > Phone...phone number > Each time a phone number is added...or changed...multiple rows are added. > If you have the following row > bo pro_cm_seq Primary effdt phone > 10 86547 Y 5/21/2004 888/555/1212 > and you add a new PRIMARY number336/555/2222 today...the following rows will > be added: > bo pro_cm_seq Primary effdt phone > 10 86547 N 7/28/2004 888/555/1212 > 10 11235 Y 7/28/2004 336/555/2222 > So now...I need the most current PRIMARY Phone number...but I have 2 records > that show as PRIMARY with 2 different numbers... I can't just do a max on > the date because the phone numbers are different. > I even went as far as islolating some of the records that showed in one pull > but not the next...if I run the sql for just that bo_id...it pulls correctly > everytime. > The table only has 2967 records in it. All fields mentioned are TEXT except > for the effdt...which is a date field.