Re: locally managed tablespace & dictionary managed tablespace

"Burt" <burtpelt_at_bellsouth.net> wrote in message news:HwTP9.2875$us1.1696_at_news.bellsouth.net...
> You seem to know a lot about this subject - lot of good FACTS right. But,
> as usual, you seem to need an attitude adjustment.
>
> Opinions don't belong anywhere in a posting, right? Wrong!
>
> Not everyone is just like you (thankfully). Some people like to hear
other's
> opinions.

I'd love to hear any opinions you have that are substantiated by anything other than vaccuousness. And I mean that honestly.

I'm not sure what you meant by "Opinions don't belong anywhere in a posting, right? Wrong!"
Are you saying that opinions are to be welcomed? Absolutely, if so.

But back it up. Or at least provide some measure as to why 23 years' experience in non-Oracle, non-DBA background warrants your opinion on a highly specifically Oracle question. Otherwise we might all just post (as a for example off the top of my head) that anyone with body hair is, IMHO, worthy of throwing off the next boat to the antipodes. At which point opinion becomes either (a) baseless wishful thinking or (b) bigotry.

And I am sure you approve of neither.

So, please. Inform me:

1. What memory management has to do with tablespace management
2. What were the results of the block dumps you performed to assess the integrity of Oracle's design for LMTs
3. Why you think the *implementation* of LMTs is so much better after 2 releases (4, if you count point releases) than the implementation of DMTs (after 6 if I count right).

Had you ended your post at the first fullstop (or period, if you prefer) I'd have had no quarrels with it at all. It's what came afterwards that caused me to doubt the veracity of your opinion. Or at least the integrity of the foundations on which it is based. You can't just post 'X' as though it counts for something, hint at why you know 'X' to be the way to go, and then fail to provide one scintilla of reasoning for that way of thinking.

Science doesn't work like that.

Regards
HJR
> "Howard J. Rogers" <howardjr2000_at_yahoo.com.au> wrote in message
> news:gZRP9.12567$jM5.35658_at_newsfeeds.bigpond.com...
> > "Bert Bear" <bertbear_at_NOSPAMbertbear.net> wrote in message
> > news:4nRP9.2529$da3.1526135006_at_newssvr11.news.prodigy.com...
> > > Andi,
> > >
> > > IMHO, LMT are the only way to go. This opinion is from 23 years of
> being
> > a
> > > operating system developer (e.g. file systems, IPC, GUI, schedulers,
> > > dispatchers, memory management, internal trace, abend processing
> > > backup/recovery applications, etc.) on three different operating
systems
> > > (IBM VM, Cray UNICOS, and IBM OS/2). The design of LMT is much better
> > than
> > > DMT (and the implementation seems to be better).
> > >
> >
> > Oh please. What exactly does 23 years' experience in memory management
buy
> > you in terms of this discussion?
> >
> > What does "the design of LMT" mean? Tell us, please, about the block
> dumps
> > you've performed to verify and ascertain the innards of said design. If
> > design is really an issue, tell us what you'd have changed, or precisely
> why
> > Oracle's design is so f***ing perfect.
> >
> > Exactly *WHY* do you think the implementation of LMT (which had quite a
> few
> > bugs in 8.1.5) is "better" than the implementation of DMT, which has
been
> > standard for well over 5 versions of Oracle, and has (practically) any
> bugs
> > it ever possessed long since ironed-out?
> >
> > Honesty. LMTs get the vote for all sorts of reasons, but design and
> > implementation are the *very* least of them.
> >
> > So what we're left with is 23 years of largely irrelevant experience and
> an
> > "IMHO". Do you not think the original poster deserves better than that?
> Like
> > reasoned argument, maybe?
> >
> > Regards
> > HJR
> >
> >
> >
> >
> >
> >
> > > http://www.bmoshier.net/bertram
> > >
> > >
> > > "andi" <andi_at_a.com> wrote in message
> > > news:auodnb$7d5$1_at_mawar.singnet.com.sg...
> > > > Hi,
> > > > newbie question.
> > > > I've read some books which mentioned about advantages of
> locally-managed
> > > > tablespace over dictionary-managed tablespace.
> > > > My question is, are there advantages of dictionary-managed
tablespace
> > over
> > > > locally-managed tablespace ?
> > > > or when should we use dictionary-managed tablespace instead of
> > > > locally-managed tablespace ?
> > > >
> > > > TIA,
> > > > Andi
> > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>
>
Received on Mon Dec 30 2002 - 05:11:49 CST