Ah, nobody mentioned anything about "ideal". Not to one-up you, but wouldn't "ideal" mean having software released without bugs? But we're not going to go yelling down that road, are we? In real life, we constantly have to make do with some percentage of certainty of this or that, and never 100%. So, by the time the Oracle RDBMS reaches "end of support", there are still some "known bugs" as well as some undiscovered ones, but far fewer than there were previously. Which is as good as it gets.On Wed, Jan 7, 2009 at 8:02 PM, Tim Gorman <tim@xxxxxxxxx> wrote:Just another perspective on all this... When a vendor ceases fixing bugs in a product, it might then finally be considered fully tested and production-ready. Using this perspective, 9iR2 became production-ready in July 2007, 10gR1 is scheduled to become production-ready this month (January 2009), and 10gR2 will become production-ready in July 2010. So, "end of support" need not mean "end of product lifecycle", but rather "end of development" and the beginning of the production lifecycle.I find this to be paricularly inacurate. It would be *ideal* if 'end of support' mean 'there are no more bugs to correct', but it actually means: 'if you hit a bug then you are on your own!' The point being: if you are truly intent on high availability, you don't go anywhere near the "latest release" of software or hardware, regardless of the latest whiz-bang feature promising reliability and availability. The problem is, once achieved, high availability is deadly boring. Personally, like most "IT professionals", I find "newer" and "latest" simply to be exciting... mea culpa... -- //www.freelists.org/webpage/oracle-l |