Oracle is not a Server, not ACID/SQL Compliant, and it is a Performance Dog

70 views
Skip to first unread message

Derek Ignatius Asirvadem

unread,
Apr 22, 2021, 8:20:31 PM4/22/21
to
Tony Andrews

> On Wednesday, 9 March 2016 at 22:07:58 UTC+11, Tony Andrews wrote:
> in the “The Null Problem is a Non-issue” thread

> I'm afraid a "conversation" with Derek S can rarely be anything else, unless you agree to be heavily patronised. Derek hates Oracle with a passion only matched by his ignorance of it. I once spent an inordinate amount of time trying to get him to explain to me why Oracle is a "row processor", a term he seems to have made up. After dozens of emails in which I had to do all the heavy work he eventually declared that he had now learned that it no longer was a "row processor", but had once been one (presumably in or before the early 1980s). Of course he refused ever to explain what he meant by "row processor" because (I inferred) he deemed me too stupid to be able to understand if he did.

I don't understand why you post confirmation of your stupidity.

It was you that explained to me that some groupies use Oracle Index Organised Tables, in which records are physically grouped by the virtue of the Composite Index, and thus the "platform" has to (must needs be) perform row processing. Are you now saying that what you, as the infamous Oracle guru, advised me of, is false ?

Second, due to the pitiful implementation, yes, many Oracle types write procedural code [row-by-row processing] instead of set processing, so in that sense, not the "platform" sense, they process rows. That is, far more than MSSQL or SYBASE or DB2 implementations write procedural code on their [actual] SQL [actual] platform.

Third, no one said that Oracle is *always* a row processor. To set that up and then attack that, is pure Straw Man. Typical of disgusting liars, as per the evidence you provide of yourself.

> I had to do all the heavy work

What "heavy work" ???

Did you have to read the manuals before answering the question ? Was not the answer resident in that marvellous, cavernous, cranium of yours ?

And with this level of attack, you expect anything more than condescension and patronising. You can't be attacking me, and begging to be treated as a small child at the same time.

The only heavy work related to our conversations, that I recall, was the benchmark that you attempted. First, off someone else's question on StackOverflow, and then your own:
__ https://stackoverflow.com/questions/4083464/design-database-relating-to-time-attribute
__ https://stackoverflow.com/questions/4375192/performance-of-different-approaches-to-time-based-data

I have no doubt that (a) coding anything in Oracle, and (b) getting Oracle to perform at all in a benchmark situation, is heavy work, that noticeably I did not have to do on the Sybase side. I don't understand why you are complaining about it, and thus advertising the non-performance of the Oracle "platform".

I did a fair amount of downgrading the table definition, and the code, at your request, due to Oracle limitations, in order to obtain a benchmark of tables+code that was reasonably comparable (no complaint, it is simple enough for me to do on the Sybase side). But again, I appreciate that that was (c) heavy work on your side.

The benchmark results seem to have been removed from SO, here it is:
__ https://www.softwaregems.com.au/Documents/Student_Resolutions/Tony%20Andrews/Benchmark%20101204.pdf

For those who are too afraid to click, or have difficulty with detail, the Oracle “performance” in the benchmark relative to Sybase, on the similar systems (even with the Sybase side at a major disadvantage) is:

1 Subquery COUNT 3.0 x slower than Sybase
3 InlineView COUNT 3.6 x slower than Sybase
3 InlineView COUNT Improved 4.8 x slower than Sybase
1 Subquery SUM Abandoned after 120 mins
3 InlineView SUM 26.4 x slower than Sybase
3 InlineView SUM Improved 36.4 x slower than Sybase

> Derek hates Oracle

Love and hate are emotions that small children have. Even the concept that hate is the opposite of love, or that if you don’t love MyLittlePony, you therefore hate it, is just adorable. Thus for you, understandably, who loves that mountain of excreta, the absence of that perverted love, is “hate”. Ok.

As evidenced, you hate the facts, especially the facts about Oracle, that I expose (a) as a matter of the normal course of my occupation, and (b) in response to your hilarious declarations. I am a Relational and Sybase/MSSQL/DB2 expert, I have never suggested that I am expert in non-commercial products and freeware, but I am quite familiar with it, due to interaction at various customer sites.

I accept that you hate me. I can live with that. Most Oracle types do.

I accept that you hate your father. Which is a terrible quandary, because if not for his existence and production, you would not exist.

> Derek hates Oracle with a passion only matched by his ignorance of it

Please provide an example of even one instance of that alleged “ignorance”. Try not to erect a Straw Man, I will expose the dishonesty and impotence in that, and incinerate you along with your offering.

Here, I will give you some material, so that you don't have to poke around the internet with a microscope. If anything I declare here false in any way, let me know. Please use scientific, technical terms, not Oracle marketing terms (eg. the use of the word "server" in relation to Oracle fails as a scientific or technical term, but it is a much-bandied; much-loved; blind-ignorance, term in the Oracle love community).

__ https://www.softwaregems.com.au/Documents/Article/Oracle%20Circus/Oracle%20vs%20Sybase.pdf

Cheers
Derek

Derek Ignatius Asirvadem

unread,
Apr 23, 2021, 3:21:07 AM4/23/21
to
Tony Andrews

> On Friday, 23 April 2021 at 10:20:31 UTC+10, Derek Ignatius Asirvadem wrote:

First, to improve the formatting in sans serif:

1 Subquery COUNT ___________ 3.0 x slower than Sybase
3 InlineView COUNT __________ 3.6 x slower than Sybase
3 InlineView COUNT Improved 4.8 x slower than Sybase
1 Subquery SUM _____________ Abandoned after 120 mins (Sybase executed in 2.16 secs)
3 InlineView SUM ____________ 26.4 x slower than Sybase
3 InlineView SUM Improved __ 36.4 x slower than Sybase

One thing I will never forget, about our interaction, is this. You had no idea how to perform a Projection in SQL. The great big fat Oracle hexpert, with thousands of posts on the various Oracle blog sites, had no clue about Projection. Populating a benchmark with 1,000 rows in a parent table and 2,000,000 rows in a transaction table, was a nightmare for you, who were doing it via cut-and-paste or pony express or two tins and a string.

Which means Projection is unknown in the Oracle love community.

But it was effortless for me, a single SELECT command. So I had to teach you Projection.

__ https://stackoverflow.com/a/4280038/484814

And what do I get in return ?
__ Ingratitude.
__ Attack.
__ a pig-teat-sucking story that misrepresented an interaction about "row-processing".
__ No mention of the failed Benchmark.
__ No mention that Oracle sucks dead bears, that it can't do various simple SQL functions, such as a SUBQUERY, without cacking itself and filling the unix system with masses of diarrhoea.

Let me guess, you wrote 42 blogs in the Oracle love shack, teaching people about Projection, as if you invented it, without mentioning me, without mentioning that you are stupid for not knowing plain and simple SQL. Oh yeah, that proves yet again, Oracle can't do plain SQL.

If I don't retaliate, it is because I am quite happy to wait for God's Justice:
__ Envy is one of the seven deadly sins
__ Impugning the known truth is a sin against the Holy Ghost (cannot be forgiven)

Cheers
Derek

Tony Andrews

unread,
Apr 25, 2021, 6:02:14 AM4/25/21
to
On Friday, 23 April 2021 at 08:21:07 UTC+1, derek.a...@gmail.com wrote:
> ...

Hello, Derek. How I have missed your charming and informative posts! Obviously a high intellect such as yours can't be expected to waste its time on a 5-year old post from a know-nothing like me. Oh wait a minute...

> Third, no one said that Oracle is *always* a row processor.

You did, on dbforums.com (no longer alive), say "Oracle is a row-processor". I had no idea what you meant by that and had a long off-line email discussion with you. My recollection is that you conceded it was not (any longer) a row-processor (but used to be) - but you never really explained what you meant by the term. Now you say "many Oracle types write procedural code [row-by-row processing] instead of set processing" which I would not deny - but that is not what you said back then (over 10 years ago?!)

> You can't be attacking me, and begging to be treated as a small child at the same time.

I merely said that you are ignorant of Oracle (which you have proved many times - and that's fine, I am ignorant of SYBASE, but I don't pretend not to be) and that you hate it (which is my interpretation of the way you talk about it). You may prefer a different word, but your dislike of Oracle drips from every sentence in which you mention it (e.g. "Oracle sucks dead bears"). Which is fine by me.

> I accept that you hate me.

I don't hate you, I pity you. You are obviously a clever man, but you seem to despise most of the rest of humanity and are quick to resort to playground insults.

> I accept that you hate your father.

What would you like me to say back? "Your mother"? Pitiful.

> Please provide an example of even one instance of that alleged “ignorance”.

My examples would be:
1. You waded in on the Oracle forum on dbforums.com to tell me I was wrong about CHAR being no more efficient than VARCHAR2 *in Oracle*. Eventually you conceded, but why butt in when you were ignorant on the topic?
2. You declared in the same discussion that Oracle's Index-Organized Tables (IOTs) were "the same" as SYSBASE's Clustered Indexes, but badly implemented, despite your knowing nothing about them.
3. You maintained that Oracle is (or later "had been") a "row processor" - it WAS one, not "some developers treated it as one" which is your position this week.
I wish that discussion on dbforums.com was still available, I used to re-read it once a year or so for a chuckle.

> So I had to teach you Projection.

And there you link to a Stack Overflow thread that I am not involved in and have never seen before. To my recollection we have never discussed Projection. Perhaps you have confused me with another idiot? You seem to meet so many.

Love,

Tony

Derek Ignatius Asirvadem

unread,
Apr 25, 2021, 6:50:26 AM4/25/21
to
> On Sunday, 25 April 2021 at 20:02:14 UTC+10, Tony Andrews wrote:
> > On Friday, 23 April 2021 at 08:21:07 UTC+1, Derek Ignatius Asirvadem wrote:
>
> quick to resort to playground insults.

You started it:

> > On Wednesday, 9 March 2016 at 22:07:58 UTC+11, Tony Andrews wrote:
<some filth>

If you can dish it out, you can suck it up. Don't dish it out and pretend to be a snowflake at the same time. Contradicting yourself so violently might not be a good idea.

Repetition of un-evidenced hearsay will not help: it only proves that you are an idiot, repeatedly.

Note that the hearsay about hearsay places you in the asylum category. The deranged desperately need validation, that is why they impose themselves on humans so much.

Hey, what's the difference between a brothel and an Oracle conference ?
The brothel has a cunning array of stunts.

Hugs and kisses
Derek

Tony Andrews

unread,
Apr 25, 2021, 7:23:17 AM4/25/21
to
On Sunday, 25 April 2021 at 11:50:26 UTC+1, derek.a...@gmail.com wrote:
> (who cares)

Time to up your medication, Derek, I think.

Tony Andrews

unread,
Apr 25, 2021, 7:31:49 AM4/25/21
to
On Sunday, 25 April 2021 at 11:50:26 UTC+1, derek.a...@gmail.com wrote:
> (who cares)

I should have remembered the saying: "Never wrestle with a pig. You just get dirty and the pig enjoys it."

Derek Ignatius Asirvadem

unread,
Apr 25, 2021, 4:47:20 PM4/25/21
to
> On Sunday, 25 April 2021 at 21:31:49 UTC+10, Tony Crybaby Andrews wrote:
>
> I should have remembered the saying: "Never wrestle with a pig. You just get dirty and the pig enjoys it."

Correct. Especially when you are avoiding the facts of reality, facts that you produced, and denying the chronology.

1. Or more precisely, facts you promised to produce but could not produce.
Oracle "benchmark" after several concessions, such as physically duplicating the "inline" tables to allow Oracle to pedal faster:

============================================
Subquery SUM() over 2M rows, result set = 1K rows
__ Sybase = executed in 2.16 secs
__ Oracle = Abandoned after 120 mins
============================================

Thus "Oracle sucks dead bears" is an evidenced fact of reality. I understand that sucking dead bears is quite acceptable to people who suck pigs.

2. Did not know plain SQL. Got taught for the "benchmark", so that it could happen without the old ladies complaining that cut-paste 2M commands is onerous. But interestingly, now forgotten. Evidently, yet again, Oracle can't.

3. It was you that entered the sow stall, devastated, motivated to make a lose-lose look like a wee-wee, and after your session with the gorgeous creature (we can only hope that she enjoyed it), started throwing mud around. I hope you enjoyed it as well, at least more than the rigorous beating you took in the evidenced facts department. But you posted filth against me in a public forum.

4. I just answered your pleas for attention, and performed my karmic duty.

5. And finally, when receiving a serve of your own filth, you then started the self-contradiction, the self-identification as a snowflake that is scared of its own mud. Awww, poor baby. Suck it up.

It is probably too simple for a poo-filled sow sucker to understand, but life is actually simple: if you don't throw your pig poop around, you won't get porcine excreta fed back to you.

========================

Now if you had one honest bone in your flabby body, if you had two cells of grey matter that were connected to each other, you would deal with the facts. Ten years have passed and the latest Oracle version is even faster at cacking itself. Give it a go:
__ https://www.softwaregems.com.au/Documents/Student_Resolutions/Tony%20Andrews/Benchmark%20101204.pdf

Oracle Non-Architecture now has even more components:
__ https://www.softwaregems.com.au/Documents/Article/Oracle%20Circus/Oracle%20vs%20Sybase.pdf

In even more pathological denial of Database Server Architecture:
https://www.softwaregems.com.au/Documents/Article/Sybase%20Architecture/Sybase%20ASE%20Architecture.pdf

In that context, it is easy to understand your denial of reality, your obsession with pig poop.

Warm cuddles
Derek

Tony Andrews

unread,
Apr 26, 2021, 4:05:24 AM4/26/21
to
On Sunday, 25 April 2021 at 21:47:20 UTC+1, derek.a...@gmail.com wrote:
> It is probably too simple for a poo-filled sow sucker to understand, but life is actually simple: if you don't throw your pig poop around, you won't get porcine excreta fed back to you.

I think this video expresses your point more eloquently:

https://youtu.be/eSrXqOI9988

Derek Ignatius Asirvadem

unread,
Apr 26, 2021, 4:37:58 AM4/26/21
to
> On Monday, 26 April 2021 at 18:05:24 UTC+10, Tony Andrews wrote:

> I think this video expresses your point more eloquently:

How adorable !

You think I would watch your animal porn !

You think you can express **me** more eloquently.  So precious !  I love it.  You can't even express you, but you think you can express me, and express me better than me. Evidently you are oblivious to that.  Oh, wait, it might be denial. I am not qualified to supply a diagnosis, and this is probably not the best place to seek it.

With that much obsessing about me, I suppose I should take it as a compliment. Thank you, you are very kind. I am nothing, just a pitiful sinner like you, except for the Graces granted by God. I could not have created that kind of Gift myself, in myself (logic).  All acknowledgement is to him alone.

Do sign your name, please. Own the filth.

Big smooch
Derek

Derek Ignatius Asirvadem

unread,
Apr 27, 2021, 3:46:52 PM4/27/21
to
Tony Andrews

After you have had your fill of sow's milk, a good sleep, and changed into a fresh nappy [diaper], perhaps you could look into the technical aspects of the thread.

> On Friday, 23 April 2021 at 10:20:31 UTC+10, Derek Ignatius Asirvadem wrote:
> > On Wednesday, 9 March 2016 at 22:07:58 UTC+11, Tony Andrews wrote:

====[ Oracle is a performance Dog ]====

1 Subquery COUNT ___________ 3.0 x slower than Sybase
3 InlineView COUNT __________ 3.6 x slower than Sybase
3 InlineView COUNT Improved 4.8 x slower than Sybase
1 Subquery SUM _____________ Abandoned after 120 mins (Sybase executed in 2.16 secs)
3 InlineView SUM ____________ 26.4 x slower than Sybase
3 InlineView SUM Improved __ 36.4 x slower than Sybase

You said //it doesn't matter that Oracle is slow with SUBQUERY() because everyone nappy user knows to code the requirement as an INLINE VIEW//, so we benchmarked both. Oopsey:
- Subquery never returned
--- perhaps the thousands of programs got tired of their never-ending chatter, and stopped talking to each other
--- same as PissGresNONsql
- InlineView is only 36 times slower.

Does the latest version of the famous self-cacking suite, of thousands of programs trying to do the same thing, do the thing it can't do, any faster ? My colleagues inform me that the latest version of the self-blocking RAC is even slower, and even better at cacking itself. They buy nappies in warehouse quantities, and enjoy warehouse prices.

Or perhaps you would like a benchmark between Oracle and PissGressNONsql. Prove that 3,000 commercial developers can produce a better result from their thousands of programs, than 10,000 academic coders spread across the planet. After all, the two of them have the same anti-philosophy, the same Stonebraker hysteria: deny the principle of a database and fabricate a subjective "reality".

Or perhaps you could deny the evidenced facts of reality, and challenge my declarations:

====[ Oracle is Not a Server ]====

====[ Oracle is Not ACID Compliant ]====

====[ Oracle is Not SQL Compliant ]====

No ?
Nothing ?
Ok, resort to attacking me personally, to form a distraction from reality. Buy nappies, because there will be consequences.

Cheers
Derek

Tony Andrews

unread,
Apr 28, 2021, 1:56:34 PM4/28/21
to
On Tuesday, 27 April 2021 at 20:46:52 UTC+1, derek.a...@gmail.com wrote:
> (stuff)

I'm really sorry old bean, but I don't have the time or inclination to go over your benchmark exercise of 10 years ago. I'm sure you have proved to your satisfaction that SYBASE is X times faster than Oracle, and I will accept that too. I have no interest in dissing SYBASE or claiming Oracle is better.

Love and peace,

Tony

Derek Ignatius Asirvadem

unread,
Apr 29, 2021, 6:13:07 AM4/29/21
to
> On Thursday, 29 April 2021 at 03:56:34 UTC+10, Tony Andrews wrote:

> I'm really sorry old bean

> Love and peace

Love and peace
Derek
Reply all
Reply to author
Forward
0 new messages