Re: Index compression vs. table compression

"Howard J. Rogers" <hjr_at_dizwell.com> wrote in message news:ct13a4$pam$1_at_news-02.connect.com.au...
> Richard Foote wrote:
>> "Howard J. Rogers" <hjr_at_dizwell.com> wrote in message
>> news:ct0s43$fhh$2_at_news-02.connect.com.au...
>>
>>>Richard Foote wrote:
>>>
>>>
>>>>There is no hidden agenda, there is no mysterious plot, there is no
>>>>personal vendetta, there is no secret monies being exchanged for
>>>>unscrupulous deeds, there is no mission to select the 10 most beautiful
>>>>male and female DBAs in the world in order to transport them to a hidden
>>>>base in outer space so that a deadly chemical agent can kill off the
>>>>rest of mankind so that the select 10 (via a FTS of course) can come
>>>>back and DBAs rule the world ...
>>>
>>>You reported me to Oracle management once for having passed on
>>>commercially sensitive information to Steve Adams. Without one shred of
>>>actual evidence, of course. And without it being even remotely true.
>>>
>>
>>
>> Hi Howard,
>>
>> Oh dear, it's now bordering on the paranoid delusional !!
>>
>> Of course I didn't report you for any such thing which you can confirm
>> with the HR department or with my supervisor at the time.
>
> Since Surinder told me unequivocally that you did, along with others, I
> shall simply stick with what I know to be the case.
>
>> That said, we both know who did report you, who can also confirm that I
>> didn't.
>>
>> Now, back to the *subject* on hand, what's all this about small lookup
>> tables being read by FTSs...
>
> No, that wasn't the subject. The word "lookup" is of your invention.
>

Howard,

Dredging up some of your personal problems from years ago whilst at Oracle, in a public forum such as this is and then suggesting that it's somehow connected to this thread, is simply bizarre. Let me re-state again and for the last time *I did not* report you to anyone. If you've secretly had a "dim view of me ever since" (which is news to me ?) then perhaps it might have helped if you discussed things with me at the time, no ?. If you want to belatedly discuss it, then simply contact me *offline*.

I suspect however, this is just a poorly directed and quite extraordinary smokescreen to divert attention from a number of technical errors you've made in this thread. I understand some people take criticism poorly but your reaction has just been ridiculous. In the past when I've agreed with something you've said or written (as I have many times), I've done so openly and happily. Great, no mention of some secret agenda then. Now that you've written a number of things I disagree with, which I believe are misleading or plain wrong and for which I have *technically explained why* I disagree (as have others such as Tom, Jonathan and Tony), all of a sudden, there's this agenda going back years !! Oh please ....

*Please*, let's stop this nonsense of secret agendas and just focus on *technical* matters. Your explanations have so far been as weak as American beer as both Tom Kyte and Tony Andrews recently (and Jonathan Lewis previously) have eloquently explained.

Let's blow away some of this smoke and see what we may find. Over there (cough, splutter), I see a large shadow on the ground. Oh goodness, it's the body of a poor horse that appears to have been beaten to death. Through the haze (cough, cough), I think I can make out it's name on it's saddle cloth: "Lookup Table". Strange name for a horse, poor thing, it appears to have taken a really nasty beating. The smoke's clearing nicely now, I can now see some evidence on what may have happened to the poor blighter. Crawled on the ground, there's a message that says:

" how is the optimiser likely to read small, useful, ***lookup tables*** ?.. er, via a FTS, probably, if they are genuinely small.", signed HJR.

Howard, please pay close attention and spot the difference. IMHO, "how is the optimiser likely to read small, useful, lookup tables ?.. er, *via an index* probably, if they are genuinely small" *especially* if as Jonathan said "they are being used for doing lookups".

I also disagree with your next quote and reply to Jonathan: "Why? A small table is always likely to be read via a FTS using CBO. Even for a single key lookup... " (which shows naivety and decided lack of understanding on this issue). Why, because that's simply not true as a *single key* lookup on ***lookup tables*** (or any table if you're using a *single key*) will return one row in most Oracle databases as generally most sites don't replicate each and every row 30 times as in your Budweiser example, meaning the overheads are "always likely" to favour an index.

Howard, *that's* my disagreement with you (and of course, all your buffer cache, cache/nocache errors).

And that is as secret an agenda as I have. If others in the newsgroup think I'm being "unethical" in what I've said, I'm sure they'll jump on me as they have on others in the past.

Sorry to everyone (especially Tom ;) for kicking poor innocent "Lookup Table" in the teeth again ...

Richard Received on Mon Jan 24 2005 - 06:08:12 CST