Re: 11gR2 Or 10gR2 on Linux x64?

From: joel garry <joel-garry_at_home.com>
Date: Fri, 16 Apr 2010 09:08:47 -0700 (PDT)
Message-ID: <d1bc23dd-e7cc-4194-9b1a-c1bb19673e97_at_g11g2000yqe.googlegroups.com>



On Apr 16, 2:06 am, Rob Burton <burton...._at_gmail.com> wrote:
> On Apr 15, 11:36 am, Jia Lu <roka..._at_gmail.com> wrote:
>
> > Hi all
>
> >  I want to use the newest 11gR2 in a new project.
> > But my boss said 11gR2 is too new. 10gR2 is the best.
>
> > I want to know if there are some fatal bug report about 11gR2 on
> > Linux ? And is 11gR2 more available with 10gR2?
>
> > Thanks.
>
> > Best
> > Lau Lu
>
> 10.2 goes out of premier support in July2010. Core functionality tends
> to improve with later releases (new features might have some bugs in
> them). I've had no problems with 11.2 so far, and as pointed out 11.2
> now has a patchset update available now.

It makes sense for, say, a conservative accounting department to be trailing edge. This is often driven by application vendor requirements and existing systems. It makes sense for new projects using newer features for DW, DSS or web applications to be fairly current. To be bleeding edge, you have to accept that you will run into perhaps fatal bugs as they get discovered, requiring expensive (in technical staff terms) solutions. The latter is a managerial decision, which may be driven by input about the usefulness of new features, and funding for working closely with Oracle in some larger organizations.

For one guy doing a new project under a conservative boss, perhaps you (Lau Lu) could point out that you could develop under the bleeding edge, but implement under the stable system if there isn't a real necessity for the newest features - and it could be a way to get a start on the upgrade curve without fully committing to it, so you can check if it will break anything existing as time allows. Sometimes the new features are useful, and as Rob says, core functionality improvements provide free benefits.

I've seen a number of places have specific budgetary requirements over, say, a five year horizon, so as hardware prices drop, they can get more powerful new hardware with new versions of everything, test them, implement, and make the old stuff testbeds/standby until the next cycle. This is much better than some places who say "if it ain't broke don't fix it," then wind up in a crises as everything is no longer supported and the "stable system" breaks anyways. There are reasons for the accounting concept of depreciation, which includes recognizing business equipment gets old and less useful. Bespoke software may add value to the business, or may be quickly surpassed by COTS. jg

--
_at_home.com is bogus.
I'll never trust Snoopy again.
http://www.signonsandiego.com/news/2010/apr/15/metlife-settles-government-payments-probe/
Received on Fri Apr 16 2010 - 11:08:47 CDT

Original text of this message