go for raw sql if you need to, but be aware that that article was written in 2007, testing only MS SQL Server behaviour .... we are in 2011, MS SQL Server 2008 is on the run (and that article could be referring only to the 2000 or 2005 version)...
Hopefully MS ingeneers have found (or will find) how to optimize different queries that lead to the same dataset in the exact same amount of time. I don't have the time to test right now, but at work I have to deal with MS SQL every day...if you want I can do some tests..... right now I wonder also if it will be faster (I'm thinking exclusively to MS SQL) to do the same query again but using a different approach: left outer join and checking NULLs. Web2py implements a way to filter the data as you want (and give you a chance to use whatever raw query you want to)... it's database's engineers work to deal with this kind of optimization (or it's on you to use raw SQL, or to pick another database at all :-P) On 6 Mag, 18:12, pbreit <[email protected]> wrote: > I didn't see EXISTS in > dal.py:http://code.google.com/p/web2py/source/browse/gluon/dal.py > > That seems like it might be a tricky optimization to program but perhaps > it's possible. > > You could also use raw sql:http://web2py.com/book/default/chapter/06#Raw-SQL
