Slashdot Mirror

a transactional storage engine that is not controlled by a hostile company

Which brings the total to at least 3 such so far: SolidDB, PBXT in addition to the well-known Falcon project.

If Oracle intended to weaken MySQL's market position by the acquisitions of Sleepycat and Innobase, they seriously miscalculated.

Great question and I am happy to address it. I'd say there are three main responses to your question:
 
First, you can't kill a GPL product by just acquiring it. If you as the owner continue to maintain it (as Oracle has done so far), nothing bad has happened. If you stop maintaining it, the community will quickly take over and make sure the software stays uptodate.
 
Secondly, I believe that the open source world has a self-healing mechanism built in. When something disappears or is at risk at disappearning, replacements quickly emerge. Completely on his own, Paul McCullagh developed the transactional PBXT storage engine for MySQL. And our partner Solid did the same. Check out their website http://www.solidtech.com./ There are other storage engines in development and under design.
 
Thirdly, as a company we are ensuring that there are strong transactional storage engines available for MySQL. MySQL Cluster is one and the Falcon project is another one. Falcon is a modern transactional storage engine built by Jim Starkey (known as the inventor of MVCC and the one who coined the term blob) on basis of technology he had developed and that we acquired. Falcon will soon come out as alfa.
 
The acquisition by Oracle of InnoDB (i.e. Innobase Oy) sent some shock vibes through the open source community a year ago, and perhaps rightly so, but I think we can be proud of what has transpired since then. As for Sleepycat (and their engine BerkeleyDB), it never was in any real use inside MySQL and after looking at the technology we decided a long time ago not to pursue that alternative.
 
Overall, there is another learning from all of this. It is a step towards increased modularity of MySQL, which in itself makes it easier for others to contribute code. We have taken an "architecture of participation" as a priority for us, and making the product modular is one important step in enabling participation. If you have good ideas on how to achieve higher modularity and enable more contributions, let us know!
 
Marten

Says here that the marketing VP quoted in TFA, Paola Lubet, used to work for Oracle.

Now I wouldn't want to read anything into that, I'm guessing there's a lot of rotation between database companies, and it's a private company held by a bevvy of investors, so Larry Ellision presumably won't be able to drop it in his shopping basket when he pops out for some milk. On the other hand the investors sound like venture capitalists (I may be wrong, corporate finance is not my field) who might be interested in any short-term returns on their investment that a large, cash-rich company could offer.

Within two clicks of the Slashdot story page, I found THIS.

Solid Information Technology Ltd http://www.solidtech.com/ Solid Information Technology Ltd is the maker of a unique data management product, SOLID Server. It is designed for robust operation in demanding environments. Its small size and ease-of-use make it ideal for deployment to web sites, various embedded systems, and other environments where databases need to operate unattended.

I'd put my money on it being Clustra any day. Why would Sun want to fool around with MySQL, or Postgres, or Ingres when they have their own HA DBMS and a truck load of developers.

Clustra was developed by Svein-Olaf Hvasshovd, Oystein Torbjornsen, and Svein Erik Bratsberg http://www.informatik.uni-trier.de/~ley/db/indices /a-tree/h/Hvasshovd:Svein=Olaf.html. Looks like Svein-Olaf - Sophus - went back to Norwegian University of Science and Technology in Trondheim. But Oystein and Svein-Erik still work for Sun at the Trondheim office.

For extra credit read:

http://research.solidtech.com/publ/drake-isas04-ha db.pdf - note location of Sun credits to this paper.

Solid!

Other than Oracle and Postgresql, I don't think any other database uses MVCC.

Postgresql is BSD style (do what you will) while SAP was released under the GPL.

Hello foog,

Remember, though, as stated in the article, much of sleepycat's position in their proprietary licensing market comes from their open-source software strategy.

Indeed. Open source as loss leader, with traditional licensing for revenue, may have possibilities. If so, this will require substantial changes to the open source/free software mythology, from a revolution model to a peaceful coexistence model. It's not compatible with the slogan that in the future, all software will be free.

It does look like dual-license strategies do significantly better than the "write free software and charge for services" model, doesn't it?

Yes, although in this case, not enough to get past the mom-n-pop level, at least not so far.

I wonder how well sleepycat does compared to other niche "embedded" database vendors like Polyhedra or Solid.

Hard to say. They're all privately held. The Polyhedra site doesn't say how big they are, though it offers a wide variety of product, service and support options. It shows no current job openings and does not have an "about the executives" page, which are not good signs. Oh wait, there's this inconspicuous note -- they were acquired in whole by Enea Data AB in March 2001 for their RTOS. Doesn't exactly sound like a high-ticket sale.

The Solidtech site says they closed a $50M venture capital round in 2000, which implies that they are probably not profitable but were seen as having the potential to become profitable. They have only one job listing at present, an account manager, which as bellwethers go is not a particularly good one.

Tim

Solid Information Technology

Sybase is far better technically than Oracle IMHO.
Heard of SQL Server? Spew. Sybase sold it to M$.

[Sybase] has a smaller footprint memory and disk wise
What's wrong with Oracle's physical and logical disk configuration? The optimal solution is 22*2 dedicated drives -> parallel read write, speedy SELECTs, good locking architecture etc...

The word according to Mr. Greenspun:
Here are the factors that I think are important in choosing an RDBMS to sit behind a Web site:
1. cost/complexity to administer
2. lock management system
3. full-text indexing option
4. maximum length of VARCHAR data type
5. support

What's the price per processor for a Sybase DBMS per year? What's the cost of a junior (12 to 18 months experience) Sybase DBA? Oracle8i is around $125,000 per processor. ~96% of the e-50 use Oracle according to their adverts. Thus, Oracle is the market leader and can name their price. Starting Oracle DBAs can earn between $60,000 to $80,000. Unfortunately the salary is much lower in good old blighty.
AFAIK, Sybase is not a player when it comes to point 3. Oracle falls down at point 4. What's the max VARCHAR on Sybase? Do me out of this job by using the relatively cheap ($2000 for 25 users), minimal administration DBMS Solid.

Regret for the past is a waste of spirit

My experience is that Solid is a very robust, easy to set up, low maintenance RDBMS (with referential integretity checks, transactions, etc.) that isn't open source. They seem to now be fancying themselves a new Oracle so I'd just stick with Oracle if you need something out of this class. Kinda sad because Solid has been around with Linux since the early days as a low cost solution, but new management has come in determined to tear the company down, it would appear...