Re: Where is Oracle’s Grid ?

Daniel Morgan <damorgan_at_x.washington.edu> wrote in message news:<1072479592.690344_at_yasure>...
> Niall Litchfield wrote:
>
> > "Daniel Morgan" <damorgan_at_x.washington.edu> wrote in message
> > news:1071847804.695870_at_yasure...
> >
> >>>>I'm with you on many of the 3rd party suppliers. But why, exactly, is
> >>>>their refusal to follow advice given numerous times by Tom Kyte,
> >>>>Jonathan Lewis, Richard Foote, etc., etc. etc. Oracle's rsponsibility.
> >>>
> >>>They CANNOT follow the advice! That IS the whole point!
> >>>They do NOT have the source code to play with, they CANNOT
> >>>change a single line of any production setup without incurring
> >>>the "lost warranty" rubbish, they don't get the slightest
> >>>support from Oracle if they dare go against the 3rd party maker.
> >>
> >>I'd buy your argument if it was impossible for anyone to build a
> >>third-party application using the Oracle database but it isn't. And I
> >>know for a fact there are a huge number of extremely successful projects
> >>built with the same starting tools used by Siebel, SAP, and PeopleSoft.
> >>
> >>Just to name one look at Amazon.com. They have built one of the world's
> >>top database implementations using the very same CDs supplied by Oracle.
> >>They too didn't have the source code to tweak. So if they could do it
> >>correctly ... and so have many others ... why is it Oracle's fault that
> >>the big three app vendors wouldn't know referential integrity if it
> >>climbed out from under their bed every morning?
> >
> >
> > Um, I think that you are arguing at cross purposes. The point isn't that
> > people don't have access to Oracle source, but that people don't have access
> > to the application source. If Amazon bought an off the shelf web site
> > package and used it to build Amazon.com (and maybe they did but that wasn't
> > my understanding) then the comparison is valid, otherwise it is irrelevant.
> > If they did what package is it does anyone know?
> >
> > 3rd party apps generally are designed to run across multiple customers with
> > different data volumes, business models, trading volumes, levels of
> > experience and software platforms. In my view it is really not surprising
> > that they are least common denominator stuff, and to a lesser extent that
> > code that belongs in the db goes into the application. Nearly all software
> > that runs on Oracle these days is produced by commercial software developers
> > and not in-house, for most of these people the challenge is to support
> > Oracle,MSSQL and DB2.
> >
> > That said, and maybe I have just been lucky, the vendors we have have
> >
> > a) generally got one or more tools that allow you to submit sql directly.
> > b) allow additions to schemas (but not 'modifications') so you can add
> > indexes, mviews etc etc.
> > c) sensible support staff who will listen to cases/demonstrations of
> > problems and agree 'workarounds' with you.
> >
> > Too often the picture is of a vendor who won't allow *anything* to be
> > changed at all, even if the payroll takes 8 days to run for 8 people. This
> > picture is I think a little misleading.
> >
>
> I'm not arguing that companies have built lousy apps using the Oracle
> database as the back end. They have also done so using DB2, Informix,
> Sybase, and MySQL as the back-end.
>
> That that good apps can be built on these databases and the fault,
> therefore is the vendors, not Oracle's.

Only partially agree. If Oracle is going to put out product that purports to have major enhancements to systems management, performance monitoring and diagnostics ( see page 18 on both http://otn.oracle.com/products/oracle9i/pdf/9i_new_features.pdf , http://otn.oracle.com/products/ias/pdf/904NF.pdf ), then it better work on what's out there, not on what might be out there in a perfectly pure Oracle world. Goes double if it claims to fix things.

jg

--
@home.com is bogus.
http://www.signonsandiego.com/news/politics/20031210-0914-nixontapes.html