Nicola

> On Sunday, 8 August 2021 at 23:52:04 UTC+10, Nicola wrote:

Dealing with the rest of that post.

> > On 2021-08-06, Derek Ignatius Asirvadem wrote:

> > 1. Read the Transact-SQL Users Guide
> > ____ at least ch 20 Transactions: Maintaining Data Consistency and Recovery
> >
> > 2. Then read my guide to the Lock Manager
> > ____ https://www.softwaregems.com.au/Documents/Article/Sybase%20Lock%20Manager/Sybase%20Lock%20Manager.pdf
> > All my Sybase docs are condensed, intended for Sybase DBAs. I have
> > just updated it, and added a bit of detail, imp[roved the clarity, so
> > as to be relevant for novices.
> >
> > Remember, this is a serious Lock Manager, not comparable to your 2PL
> > filth, which has to be asserted because you guys position commercial
> > SQL Platform Lock Managers as your "2PL" filth, and insist that we
> > have your insane problems. It is so mature and secure, so brilliant
> > in architecture, that it has not changed since 1984. Extended, yes
> > (eg. to handle new Lock Types to support SAP files, eg. add row locks;
> > etc), but changed, no. So read these docs with a fresh mind, to not
> > take your academic baggage with you.
>
> I have read a couple of documents (the T-SQL guide and the Locking and
> Concurrency Control manual).

You are best advised to read the T-SQL UG *FIRST*, because it is an explanation, the P&T series is reference, not explanation, so read that *SECOND*, after you understand the explanation. Otherwise you will continue finding “faults” and “problems” where there are none, which is your documented occupation.

> My takeaways:
>
> - yes, it is a serious lock manager (I did not expect anything less from
> a high-end commercial product).
>
> - I did not find any concept that you would not find in a database
> systems' textbook (and no, the Alice's book is not such a textbook).

You dishonestly fail to mention the fact that the textbooks were written AFTER Britton-Lee, AFTER Sybase. Your point is null.

The other ramshackle book that you linked is no better, no worse, than the Alice book. They fail totally in teaching science, they succeed well in teaching how to suck a pigs rectum, which is professed to be “science”. Every link you propose, presented as “knowledge”, has proved to to be sows milk on top of a fragment of science.

> - Call it what you like, but ASE/Sybase uses what is known as "rigorous
> 2PL" to implement repeatable read and serializable:
>
> https://help.sap.com/viewer/a08646e0736e4b4b968705079db4c5f5/16.0.3.7/en-US/a8ea3fd4bc2b1014b6569e800f6bba42.html.

There is no mention of a “2PL” or “rigorous 2PL” in that.

I don’t call it “2PL”, I detest calling something what it is not, that is fraudulent, anti-science. The term is something the academics created, it sort of explains their two phases (they do have a hexpansion phase and a contradiction phase). Calling their filth, as implemented to provide some concurrency control in the MV-non-CC herd of programs, 2PL, is correct.

Calling the Sybase; DB2; MS, Lock Manger “2PL” is false. It will prevent you from finding out what a real Lock Manager is. Further, as explained in detail earlier in this thread, and again in the Transaction Sanity doc, it is a Straw Man, erected by anti-scientists, to demean and discredit something they (as evidenced) cannot comprehend.

The point I made, which stands unaffected, is that it is idiotic for anyone to call something what it is not. Repeating, it is the anti-science academics that say that commercial SQL Platforms are “2PL”
____ WHICH IS FALSE
and as such, as a false “2PL”, that we somehow have the same “problems” that their idiotic 2PL lock managers have. It is a Straw Man.

You too have done that, just look at your questions. You are still doing it. You are the one doing this “calling it what you like” stupidity. You can stop making a fool of yourself any time you like.

If you genuinely understand what you are saying NOW, you would realise that your questions pertain to the poor implementation, your 2PL, and not to the real 1PL lock managers.

> - Call it what you like, but ASE/Sybase uses what is known as "rigorous
> 2PL" to implement repeatable read and serializable:

Go on, name the two phases in the Sybase Lock Manager.

If you hide behind “what is known as”, then take responsibility for the “what is known as”, because it is you asylum screechers that have written the “literature” such that “what is known as” is what is known as.

I just do not play your stupid game of calling things what they are not.

> "Applying exclusive locks [...] until the end of the transaction.
> Applying shared locks [...] until the end of the transaction".
> Textbook definition of rigorous 2PL.

Except that you are fraudulent in applying that to a non-2PL server.
And that the textbook was written after Sybase.

What you have is a fat hairy sow, with a penchant for academics. Sybase is a horse. You say that a pig is the same as a horse. I say, pig poop. You say a macho hairy boar can do anything a horse can do, especially when it is rigorous. I say, fiddlesticks, fresh warm pig poop, straight from the sow. More like rigor mortis.

> - It uses index-locking to prevent phantoms. Again, no surprise and
> pretty much standard textbook material.

Except for the fact that Sybase came first, and the texbooks came second, therefore the inference is false.

> - There is a section dealing exactly with the question I have posed:
> "Locking for Select Queries at Isolation Level 1"
>
> https://help.sap.com/viewer/a08646e0736e4b4b968705079db4c5f5/16.0.3.7/en-US/a8eb04a3bc2b1014bef8884d8400b0ab.html

I don’t see how that has anything to do with your questions, which were your particular speculations about how low level operations could be, should be, wannabe, please be, problematic. The usual insanity of imposing the problems of the insane, onto the sane. It is such an obsession with you, that even after it is pointed out, you cannot stop.

I tried to explain the operation of the Sybase 1PL Lock Manager, and failed.. Daniel did a better job, and it appeared that you accepted it. Probably not. Now you have found a relevant page is the manual. Excellent.

The section is simply a definition re how a particular sequence of commands works. Not about “exactly your question”.

> Btw, with a mention of how that affects joins.

Yes, of course. So what. Even our joins are superior to the way you guys speculate about joins forty years after we had them.

The Sybase manuals are not as good as they were before the acquisition by SAP. If I were to give that definition, I would include the fact that it intent-locks UP the [SELECTed] tree as well, ie. the PK page/row in the parent table that is referenced by the FK in the child table which is named in the SELECT command. But hey, that is only a reference manual. And you would not understand it anyway, because it is relevant only to Relational databases.

Do you know wtf the difference is, between APL and DPL/DRL lock schemes (tables). Sure, you can state the words, but do you know the RELEVANCE. No. Do you know why there is a difference at the level described in that section ? No. But somehow, you say, it applies to your question, which was before you read that manual.

> That, plus this (which
> is about SQL Server and has some inaccuracies, but overall I think it
> is relevant):
>
> https://sqlperformance.com/2014/04/t-sql-queries/the-read-committed-isolation-level
>
> makes me conclude that in general you do not have statement-level
> consistency at read committed in SQL Server or ASE.

God help me.

Destroyed, in my previous post. In sum:

1. It is a sign of desperation, when an academic cites a non-academic blog post. Usually it is the other way around.
2. It is stupid to expect a COUNT on a large active table in an online shared database, to be accurate in reference to any fixed notion, particularly when the fixed notion does not exist, has never existed. Humans who have not been indoctrinated into schizophrenia are not that stupid.
3. It is Oedipus stupid to expect such a COUNT at READ COMMITTED to be executed as REPEATABLE READ, when the manual clearly states that you can obtain such a COUNT at REPEATABLE READ or SERIALIZABLE.
3. Last but not least, anyone with actual experience on a genuine SQL Platform can obtain such a meaningless but accurate count without having to plod through each and every page in the table, SQL-Standard-wise. A true instantaneous COUNT, and a true point in time.

----

> ASE is a fine implementation (*), based on concepts that have been very
> well known in the academic community for a long time

Lies, more filthy lies. “Based on” is false. The father is not based on the son.

The historical evidenced facts are:
___ Britton-Lee & Sybase
___ Lock Manager with 66 lock types (at 15.7, even more in 16, without changing the Lock Manager)
__ academia pushed the Stonefreaker MV-non-CC as “MVCC”. It never worked in Ingres, it has not worked in PigPoopGres. It never will. They added their 2PL on top on their MV-non-CC, to get it to work at all. It still doesn’t work.
__ They only know their 2PL.
__ Sybase, DB2, and MS do not have 2PL
__ But they, and you here, falsely re-frame a Lock Manager as your hysterically stupid 2PL.

Therefore “based on” is disgusting filth.

You can call it what you like. You can reframe and redefine terms until the cows come home. Reality does not care what you call it, it is not affected by what you call it.

> (not to say that
> they are obsolete! On the contrary!). What's not in the textbooks is the
> specific implementation details and system-dependent guidelines that
> a manual is expected to provide. Granted, the devil's in the details.

Which, as evidenced, is beyond the ken of those who write the textbooks and academic papers. They cannot even copy-paste, they have to deny the reality of forty years, and re-invent the wheel from scratch.

Forty years, and counting.

> But good graduate students would have no problems grasping such details
> (or those of any other system), capitalizing on their academic baggage.

Sure. Like when I give them a task to obtain a COUNT() from a table, and instead they give me an hysterical slew of “reasons” why COUNT() doesn’t count, or that there is no point in time that count is accurate, or that count must be accurate to a point in time that does not exist. I enter the few keystrokes required into their session, on their computer, without hitting RETURN, I ask them to watch the second hand on their wrist watch, and to call out the second they wish to have the COUNT for, and enter RETURN. Most hackademicks lose control of their bowels. On two occasions they quit overnight and did not even have the courtesy to have a resignation meeting in person. Another committed suicide a month later.

Sure. It is only when the rubber hits the road, when they get a job in the real world, that their baggage is exposed for what it is, and those not totally indoctrinated get rid of it. Those attached to said baggage, have difficulty dealing with reality, that truth is simple, it is only falsity that is complex, they don’t last three months. They get very hurt when we laugh at their voluminous speculations of disaster instead of writing code.

Sure. They have “no problems grasping such details”, but they are impotent in determining a resolution, they argue endlessly about the “grades” details. Meanwhile I give the capable developers who gratefully do not try to “grasp such details” a Template, and say do this. And they do it. Hint: check out how Dan has progressed in his thread, vs how you argue endlessly, with utter futility, and you still cannot grasp the even the principles, while labouring over details.

Make no mistake, this is a life-or-death struggle, that ends in death. Suicide is irrefutable evidence of mental illness. The problem these days is, now that mental illness is being “normalised”, now that schizophrenia is being taught at universities (your various arguments, eg. how a COUNT() could be sabotaged to fail ... and therefore it applies to ALL Statements), mental illness is not perceived for what it is, it goes untreated, and we find out after the person has killed themselves.

I have a friend, 35yo and very intelligent on the IQ scale, hopelessly insane on the intelligence scale. PhD in AI. Three jobs in three years, and it is always the employers fault, never hers. Under tremendous pressure of her own making. I tried to help her because I was expert in a certain vertical in one of her jobs. But she was entrenched in RFS+”Ontology”+”Description Logics”, while begging for a real Relational database (NO ”Ontology”, NO ”Description Logics”). She has a 9yo son. Divorced. Hates men except when she wants sex, and then she turns into a siren. Her sister and I are the only ones who address her denial of reality (schizophrenia), who are trying to prevent the predictable.

Make no mistake:
__ Truth = Reality = life
__ Falsity = Rejection of Reality = death

__ Sanity = conforming the intellect to Reality
__ Insanity = desperately try to conform reality to the mind, which after a lifetime of failure, ends in suicide.

Academics foster death. It is evil.

> (*) Known to the academics. E.g., some time in the '90s, Sybase was used
> for lab exercises at Stanford.

1. And notably not taught. Academic filth taught instead. Same as academics pushing ERD, never mentioning IDFE1X for FORTY YEARS.

2. “Known to academics” is a general statement, because it implies academics generally, as such, it is a pitiful, bald-faced lie.

3. The truth is, yes, it was at Stanford. No, it was squashed by the Stonebraker groupies at Berkeley, who were pushing their MV-non-CC barrow. These decisions are political, not scientific.

4. If it had succeeded at all, they would have been using Sybase since then.

5. Unfortunately they did not study the Sybase Lock Manager, or even copy-paste the manual. As evidenced, they remained vociferously ignorant∑ and erected a Straw Man by descriing it as their primitive 2PL.

Cheers
Derek