On Thu, 11 Aug 2011 20:15:43 +0000 (GMT) Colin Wright <cdwri...@gwi.net> wrote:
> It looks like it could be really slow in a large data set. If this > is done a lot you'd want an index on kmmSplits(accountid, postDate) > (or kmmSplits(accountid, txType, postDate) if it does lots of > type-based searches and N is uncommon). Unfortunately, 'N' is the most common. The alternative is 'S' which is used only with template schedule transactions, not 'real' ones. There is already an index on AccountId + txType; you can use 'Tools/Generate database SQL' in KMM to see the schema. Please remember if you create indexes or views of your own not to name them beginning with the three characters 'kmm' which are reserved for 'official' tables, etc. -- Cheers, Tonyb _______________________________________________ KMyMoney-devel mailing list KMyMoney-devel@kde.org https://mail.kde.org/mailman/listinfo/kmymoney-devel
