> > hashtext >
I believe it's considered an internal function, and not one to rely > on. If you search the archives, you will find reports around a version > upgrade when it changed the output for a certain input and thus broke > yes. I did find that discussion, and a blog post by Peter E. about this breakage. BUT... http://kaiv.wordpress.com/2007/07/23/decreasing-the-index-size-on-wide-columns/ then there is skytools ... using hashtext for various things, and the report: http://postgresql.1045698.n5.nabble.com/Which-is-faster-md5-or-hashtext-td3251399.html describing it is fact roughly 40% faster then md5 AND is an integer, so select sum(hastext(t.*::text)) as signature from datable t as a fast way of comparing 2 replicated tables is really something to consider.... > conclusion was that it's not documented because it's internal and > you're not supposed to use/rely on it. > > My impression is that people are allready using it, relying their sharding on it, even building indexes on it. So... I suggest we start documenting it. Even if there was a recommendation not to use it, people will get hurt anyway if their application rely on it and it breaks. Could I persuade? Harald -- Harald Armin Massa www.2ndQuadrant.d <http://www.2ndquadrant.com/>e PostgreSQL Training, Services and Support 2ndQuadrant Deutschland GmbH GF: Harald Armin Massa Amtsgericht Stuttgart, HRB 736399
