From: BD

To: Editor

Date: 06/30/2003

 

You might be interested in Prevayler. It is basically a persistence engine that masquerades itself as a DBMS. At least that is what they want to replace. I know that the world is full of such products, but this one has such a ludicrous "skeptical FAQ" which is just waiting to be picked apart. And they are slowly gaining in popularity in the Java community. I hope this is just a fad, though.

 

Some "pearls of wisdom":

- a rather curious notion of transactions

- querying means navigating through (object) graphs

- a logical fallacy

 

On the other hand, they have the decency to (sometimes) say that their product is actually not a DBMS. Well, they actually wrote, "is not a database” but they probably thought of the former.

 

 

From: Fabian Pascal

To: BD

 

Thanks. I am swimming in this kind of stuff. There are so many ulcers I can accommodate.

 

From: AC

To: Editor

Date: 07/24/2003

 

You are frequently fond of referring to a new technology that will truly implement RDBMS theory.  I cam across this product the other day, and I found its combative tone intriguing and more than a little annoying:

 

Freedom

 

With Prevalence we are finally free to create true object servers and use objects the way they were intended all along.

 

We are able to use any algorithm, data-structure and query language we please. We are no longer constrained to the ones provided by database and application servers which must run on disk data-blocks.

 

We believe the whole OO community is finally free to recover from the atrophy caused by database and application server restraints. We no longer have to distort and maim our object models to satisfy their limitations.

 

We no longer have DBAs imposing us database layout restrictions. We have freed them to do something more useful.

 

We have set fire to the table-models on our walls. We have deleted our database creation scripts. We no longer have to keep them updated.

 

We no longer have to license, install, configure and maintain a database and application server every single time we want to develop, demonstrate or deploy our systems for any of our clients. Give us a Java VM and we are good to go.

 

I think the most interesting thing about this site is that there is almost no mention of tedious things like data integrity.  By default there is no rollback support - their assumption is that programmers will check all possible conditions that might cause a transaction to fail, so there is no need to recover data inconsistencies from a failed transaction.

 

Of course, any critique we might care to offer will simply be written off as

"an emotional attachment to RDBMS".

 

 

From: Fabian Pascal

To: AC

 

This should not surprise anybody; it's the logical conclusion in an anti-intellectual system that punishes knowledge and rewards ignorance. The completely ignorant are free of any constraints in their own undeveloped minds and the whole society--which prides itself on efficiency and progress--pays the price. The industry has been regressing back to the stone age of programs and files for quite a while now.

 

 

From: M

To: Editor

Date: 09/24/2003

 

I'm sure you've seen it, but a colleague pointed that the Prevayler site out to me and I about fell out of my chair! There's also a whole lot of misinformation going on there.

 

When's that new technology for the TRDBMS coming out?  Boy do we need it!

 

 

From: Fabian Pascal

To: M

 

Those things come out of the woodwork in tons and there is neither time, nor any point to bother. One must be very selective in responding.

The problem is systemic and cultural: it is an abysmal failure of the US educational system much beyond computers and databases and there is nothing you or I can do about it.

 

Regarding the TransRelational™ Model, DK and am not very hopeful in the near future. After all, it's the right thing, so why should you expect it'll fly? And what makes you think it will be appreciated amidst ignorance?

 

 

From: PM

To: Editor

 

I love your site and I know it has definitely helped me to consider database theories and implementations more carefully and in terms of the underlying theoretic models.

 

I don't suppose you've heard of Prevayler? Its Wiki-based discussion seems like a breeding ground for many fallacies and incorrect notions concerning the necessity of effective query capabilities and ACIDity.

 

 

From: Fabian Pascal

To: PM

 

I got several emails the last few months on Prevayler. Lost cause. Suggest those with some brain should simply ignore it.

 

 

From: PM

 

I'm not quite ready to give up on the innovation in the realm of databases that are tightly bound with object systems (because object-oriented programming is so common and more than ever needs strong database management capabilities), but I think the research needs to be based on sound foundational theory (which I too am a believer in that theory being relational.)

 

Thanks again for being a source of reason amidst the chaos.

 

 

From: Fabian Pascal

 

The fact that something is common does not make it right, in fact, quite the contrary. OO is at best engineering, not science, and has a lot to answer for.

 

From: Klaus Wuestefeld

To: Editor

Date: 09/28/2003

 

I'm the author of Prevayler. Do you know it?

 

Where you read "objects", you could also as easily read "data"

 

I would like to know what you think of it.

 

 

From: Fabian Pascal

To: Klaus Wuestefeld

 

No, I do not know it. I am staying away from products.

It looks like it attracts a lot of people who reveal huge ignorance about data fundamentals and that's not very promising.

 

 

From: Klaus Wuestefeld 

 

Do you have an example?

 

 

From: Fabian Pascal

 

A few readers sent me URLs I recall skimming thru them and decided they were a waste of time, too much crap to bother with. I still have the URLs but don't have time to dig into old mail to locate them

 

From: Klaus Wuestefeld

 

More important than the product is the concept. It is really VERY simple.

 

The FAQ might interest you.

 

If you find anything initially absurd, please run it by me.

 

The idea is so simple, many database people actually reject and get angry at it.

 

 

From: Fabian Pascal

 

Unless it has a sound theoretical foundation, simple means simplistic. I suspect the latter.

 

 

From: Klaus Wuestefeld

 

It is provably correct.

 

 

From: Fabian Pascal

 

I have no idea what that means.

What did you replace predicate logic with?

 

From: Klaus Wuestefeld

 

A person on the Prevayler list mentioned he could write a correctness proof for the prevalence concept. I don't think I could, though.

 

Stop suspecting.

 

Please answer: If you had enough RAM to hold all the data in a particular system, and a completely fault-tolerant machine, would you still use a database in that system?

 

 

From: Fabian Pascal

 

Stop wasting my time.

You are asking irrelevant questions that reveal you DK what a database and DBMS are.
Before you embark on designing products you should first learn the fundamentals.

 

 

From: Klaus Wuestefeld

 

I have developed two object-relational persistence layers in Smalltalk and one in Java, before I wrote Prevayler.

 

You are just being evasive. I wonder why.

 

Sorry for wasting your time.

 

 

From: Fabian Pascal

 

So what? The question was whether you know what databases and DBMSs are. Your comments reveal you do not. That validates my argument that practitioners get involved without knowledge of fundamentals. You are not different. Experience is not sufficient without knowledge.

You just don't see how ridiculous your question is. That's a level of discourse I do not have time for. There is a limit.

 

 

From: JC

To: Editor

Date:11/10/2003

 

Read this nonsense: http://www.wayner.org/books/rambo

 

"Throw Away Your Database" is the rallying cry of the Prevayler group (prevayler.org). Their open source project offers all of the protection of a database without any of the overhead. They even offer some benchmarks that show their solution running 9000 times faster than one using Oracle.

 

This talk will peel back the layers on this fascinating, ultra-lightweight solution and show when it works and when it's just a little too lightweight. Will you throw away your database? Will yours run 9000 times faster? It depends upon your application.

 

Bio: Peter Wayner is the author of the Java RAMBO Manifesto a light-hearted look at how RAM-based objects can speed your applications and simplify your code.

 

 

Ed. Comment: I rest my case.

 

 

Posted 11/28/03

 

 

 

[ABOUT] [QUOTES] [LINKS]