Re: Is it just me?

"Mark Townsend" <markbtownsend_at_comcast.net> wrote in message news:SkGVb.126082$U%5.608570_at_attbi_s03...
> Howard J. Rogers wrote:
> > Or am I not alone in feeling rather underwhelmed by the 10g new features
> > document? A tidy bit of bug-fixing, some nice but relatively minor new
> > capabilities, and -of course- the ability to do Blast queries (handy
when
> > I'm trawling through my own genome, which naturally I do on a daily
basis.
> > Not). Oh, and all that grid hoo-hah too, I suppose -which maybe 20% of
the
> > installed base might find useful some day.
> >
> > Nothing gob-smackingly 'wow I wish I had that right now' though (which
> > automatic undo, flashback and data guard did for me with 9i, and locally
> > managed tablespaces did for me in 8i). Which is all a bit of a downer,
> > frankly.
> >
> > Niggles, too: why has streams (to name but one feature) been "enhanced"
to
> > support LONGs and LONG RAWs?? I thought those data types were deprecated
> > ages ago, and had a shelf-life measured in months. Oh well: that'll
teach
> > me.
> >
> > Yours musingly,
> > HJR
>
>
>
> Hmm - I'd be interested in getting feedback on this. There are 360 odd
> new features, and this has been our biggest release since Oracle7. If
> you liked automated undo management, then automated storage management
> should be a highlight.

Agreed. That's the one definite new feature where I started to get the "ooh" feeling. It quickly wore off, though. Maybe it was the new amino acid search facilities that did that to me!

>Flashback has been significantly enhanced.

Also agreed, but 'enhancement' is what I was getting at: I don't get very excited at 'features' such as 'now you can partition IOTs in X ways' (when really you should have been able to do that the day IOTs and partitioning were invented). Similarly, "now flashback can do Y" feels just like "now it can do what it should have done from day 1". I'm not knocking that sort of thing -as I put it originally "a tidy bit of bug fixing" -maybe I should have said "feature filling-out" too. However, incremental feature enhancement isn't really (in my cynical old book) 'new features'. It's all welcome, of course, and I wasn't intending to say the product was bad -but it all makes me feel this should have been 9i Release 3, not 10g.

> Perhaps it's the document ?

Could be. It's as dull as ditchwater!

>We autogenerated the New Features Document
> this year from the Project/Feature database so that it just lists the
> new features, and there is very little positioning or rah-rah in it.
> Perhaps this was a mistake ?

Well, I'm surprised the marketing team let you get away with it! I actually prefer bare facts to marketing hype, though, and I was pleasantly surprised that the grid stuff got listed just the same as everything else with no rah-rah at all. So no, I don't think the approach was wrong. Full marks, indeed, for getting techie accuracy instead of over-blown sales spiel. I do think you can be technically accurate and amusing/interesting at the same time though!

I don't know quite how to put it, I suppose, but I'll give it a whirl. When 9i came out, it was very hard to teach it when most students were still working in 8i, because you basically had to teach rollback/undo in two completely different ways. When 8i came out, it was difficult to teach students more likely to use 8.0, because you had to teach the 'create tablespace' chapter effectively twice, once explaining dictionary managed stuff, and once the local management concept. Again, a startlingly new way of doing things, and twice as much material to teach as before. I guess that's the (subconscious) measure I use for whether something is startlingly new or just incremental. But at first blush, I can't see I'll have to teach 10g astonishingly differently from 9i. Even all that self-managing advisory stuff is just an extension of the advisories we already had in 9i. Bigfile tablespaces don't render "smallfile" ones extinct, and accordingly would merit an extra sentence or two of explanation, but not a complete re-teach. I like some of the new stuff about RMAN -but it's still clearly RMAN under the hood. You get the idea.

Maybe my expectations were too high. I'd be interested in hearing from beta testers and others whether they find it all a radically new experience, or just a fixed/more convenient/nice-to-have extension of the old. I suppose the bottom line is: did it really merit the new version number?

Regards
HJR

-- 
--------------------------------------------
Oracle Insights: www.dizwell.com
--------------------------------------------