All, > > I'm thinking the multi-table joins performed twice for each search > > will make more of a hit in performance (and maintenance) than > > writing two rows for each relationship and then always being able to > > search only once. > > I've used code like that successfully for 4 years. I'm not sure if > that's even technically denormalization, but the savings in read time > definitely outweigh the extra writes and storage.
In many modern web application, relational pedantic have been fading. Denormalization is often a good thing, and very large sites are in fact highly denormalized. This is due to the type of data they need to store (friend relationships is a classic one), and the level of throughput they need to achieve. The importance of transactions faded when MySQL become the dominant database on the web, and now even relationships are fading (sounds like a soap opera, I know). Instead, things like document based DBs and memcached is where it's at. The goal really is key/value pairs with insane performance because there's no overhead to manage that "old clunky SQL" - in fact, these are often considered schemaless databases. We have a couple of presentations coming up about this in the fall - check out the calendar on www.nyphp.org Some useful links: http://en.wikipedia.org/wiki/Document-oriented_database http://labs.google.com/papers/bigtable.html http://aws.amazon.com/simpledb/ "The times they are a changing..." (yet somehow that phrase always hangs on :) H _______________________________________________ New York PHP User Group Community Talk Mailing List http://lists.nyphp.org/mailman/listinfo/talk http://www.nyphp.org/show_participation.php
