Re: ** latest stable oracle 10G client

From: Greg Rahn <greg_at_structureddata.org>
Date: Sat, 7 Feb 2009 22:20:43 -0800
Message-ID: <a9c093440902072220w6944504aofbd63d3f406fcf37_at_mail.gmail.com>



I always find these "which release should I be on" conversations interesting. That being said, I think there is no right or wrong answer. I think 11gR1 offers some great new (simple) features and for a green field application, I would always go with the latest release. The main reason for that is, you have saved your teams from having to plan and test a db upgrade(s). You might think that a newer release has more bugs, but the counter argument to that is new releases have more bug fixes in as well, right? It's not really about numbers anyway, it's about which bugs *you* hit. If you do hit a bug in the latest release, if it can be fixed, you are much more likely to get a one-off patch (and probably sooner) than if it was an older release. The reason for this is that bugs are fixed in the main code line and then backported to the various versions. The further away from the main code line a release is, the less likely that fix can be backported as there may be code dependencies, changes that make it unfeasible.

As a matter of procedure, the Real-World Performance Group always uses the latest GA release and patchset. This is a way for us to get familiar with new features, etc. and to find and file bugs before most customers start using it. I don't think that 11g has had any more or less bugs than any other release I've worked with, but it certainly does have core features in it that I would surely miss if I had to go back to 10gR2.

I also think release time line should be considered. The recent list looks like:

10.2.0.3 ->
11.1.0.6 ->
10.2.0.4 ->
11.1.0.7 ->

??

Take that for what it is worth...

-- 
Regards,
Greg Rahn
http://structureddata.org
--
http://www.freelists.org/webpage/oracle-l
Received on Sun Feb 08 2009 - 00:20:43 CST

Original text of this message