Re: What is a database?
Date: Fri, 28 Feb 2014 06:37:05 -0800 (PST)
Message-ID: <41fc669c-6f57-419c-af9b-c67ef99755da_at_googlegroups.com>
> On Friday, 28 February 2014 18:05:15 UTC+11, Erwin..._at_ikan.be wrote:
I see you are up to your old tricks again. History has not changed in centuries. All great lies and frauds have a bit of truth attached to them.
Normally, I would ignore that sort of post, but the misrepresentation is so gross, the lies twisted with the truth so horrible, for the benefit of the readership, I must disentangle it.
The context is, normal humans, can read, have read, and have been using ACID Transactions since the 1960's, without any drama whatsoever. High Concurrency does not need to be coded, because it is the purposeful result of understanding same and implementing it. Since it is a design methodology, not an aspect of the platform, High Concurrency can be obtained on any platform, both the genuine SQLs, and the freeware that is fraudulently proposed as SQLs.
Now for the spaghetti:
> MVCC [...] is only for retards who don't know how to [write] transactions properly.
True.
Ones who are scared of them. Ones who do not understand ACID, and therefore drop acid. Then the retards have to create a substitute. Predictably, that substitute, produced by retards who don't understand transactions (don't forget the orgy of citations), the likes of Snotgrass the wesearch bellow at MicroShifty Humaniversity, is an abortion:
MVCC
> bring[s] the system down on its knees.
Contrary to the retarded "science" that "proved" it wouldn't.
Proving, yet again, the "science" and the purveyors, to be retarded.
Proving, yet again, that the citations are merely an orgy that only the retarded gyrate in. They are of no value whatsoever. Ok, they have the same value as personal lubricant.
Predictably, as retards, they cannot observe the physical universe accurately or completely, so they somehow miss the system on its knees. Determining the actual cause of it is totally beyond them.
Predictably, they obsess about their "creation" (abortion to the humans), and they dream up demented impoofments ...
> ... optimistic concurrency management ...
They give more "lectures" and "conferences", and write more "scientific" thrillers and "papers". More citations and self-stroking and group hugs. They haven't read Krüger and Dunning, or standards documents.
Predictably, they will produce MVCC squared. We don't have to wait five years for the "proof" to assault the universe.
The retards cannot understand [I]solation, so they ("hell, I can invent anything from scratch") code it, in a darling attempt to implement it, to enforce their fantasies on the real physical world, that has thus far avoided suffering same.
True. Can't argue with that.
But there is a pre-requisite. You must understand ACID Transactions. That includes the [I] for Isolation. In the database sense, not the mental illness sense. (More, later.)
Standards are published by authorities, and are available free. Therefore it (the standards and the method to avoid the systems on its knees) is available freely, to anyone with two fingers and a bit of tissue between them, who can type "google ...", and read the results returned.
Retards are fearful of authorities, they are their own authority ("hell, I can invent anything from scratch"). In retards, that issue is not very grey.
> ... Derek['s] [OLTP] standards ...
True, I have invented a few things, but I can't take credit for something that I did not write. The standards existed before I ever touched a computer, before I ever wrote anything of value in the industry. I first saw them in 1976, used with effect, and misused, with negative effect, on IBM Mainframes. The CICS TCP boys deserve all the credit.
From 1981 to 1987, I worked for a great DBMS platform supplier, Cincom (fifth largest then, unheard of now). I was on the minicomputer DBMS team, which was early days, but 100% online vs the mainframes, which were 100% batch with an online front end.
Ok, I did translate the CICS OLTP Standards for implementation in fully online systems. Of course, I wrote part of the DBMS code, I was fully aware of contention issues, both in the DBMS, and in the user code, so I was suited to do that.
Ok, I did write some education for the consultants to provide to our customers, such that they never wrote a deadlock; never experienced contention problems; always enjoyed high concurrency, etc. Of course that included a full review of ACID Transactions, yet another Standard, that I did not write.
As the industry evolved and I moved on from high-end DBMS to high-end RDBMS platforms, Sybase and DB2 in particular, I rendered the OLTP Standard and the education for those platforms. The problem space; the contention; the resolution; has not changed in fifty years. The science; the standards; ACID Transactions have not changed. The success and the guarantees can be repeated on any platform.
> > If you just follow [...] [OLTP] standards, you won't ever need any of that crap, and you'll certainly never bring the system down on its knees.
> True. Can't argue with that.
> But there is a pre-requisite. You must understand ACID Transactions. That includes the [I] for Isolation.
If you do not, that can be obtained from any company that supplies formal education on the subject. Sure, I am one of them.
If you do not have the capacity to understand ACID Transactions, you are beyond help. I try to ensure that the customer does not contaminate any education that they ask me to supply, but ultimately it is their retard, their problem.
Oh, wait. The retards. Those who cannot read anything except papers that they orgiastically cite, they are going to redefine the problem space, from scratch; redefine transactions; redefine the solution. And implement it. In the platform.
Um, the problem is not in the platform. Therefore no amount of coding in the platform is going to resolve the problem. The platform attempting to do so, will bring the host system to its knees. Predictable.
Idiots.
Please stop shifting about. Received on Fri Feb 28 2014 - 15:37:05 CET