In a personal opinion I dont think Oracle will be able to dominate the Db world.

Here is an interesting reports regarding DB Revenue in the latest versions from Both

http://www.crn.com/software/204300352

Umm, SQL Server on Linux? I don’t think so.

First, you can compare SQL Server & Oracle. As a proof, the Transaction Processing Performance Council did it, and it is an independent organization.

Second, SQL Server can run on any platform as well. You’re wrong about this.

Third, in terms of features available and ability to store large data (from 20To to 100To with SQL Server 2008), SQL Server competes with Oracle frontally.

Fourth, LINQ to SQL will be of huge value to integrate queries in C#, for the Visual Studio developer community.

Last, I haven’t said Oracle is a bad database. It’s on the contrary an excellent product. SQL Server just happens to gain momentum, and nobody can neglect that the product has improved a lot until reaching and maybe outperforming Oracle with its 2005 & 2008 versions.

oracle is a different league. sql server is not worthy enough to be compared to the mighty oracle

t-sql is the worst (non programming )language i have seen.

the programming, exception handling feature of oracle is absolutely great. sql server is trying hard to implement in sql server 2005

sql server is no match for feature rich oracle which can run in any platform

there are many feature in oracle which microsoft is trying to implement and the syntax in sql server is miserable whereas syntax in oracle is so simple to implement and you can write complex procedure in oracle which sql server will never be able to copy

Oracle sells an excellent database, everybody knows that. So do MySQL, Sybase & IBM. My point is that Microsoft, with SQL Server, has improved a great deal over years. And now, the chance is high that SQL Server has everything it takes to lead the market in a few years from now.

I hope I’m making myself clearer. I apologize to you, Ghislain, & Xu for appearing (maybe) a little arrogant on the database issue.

I’m a bit surprised that you fell into this “A RDBMS outperforms X times An another RBMS”-propaganda. As Ghislain pointed out, this depends on so much on the application and benchmark objectives that they’re quite useless statements. I’d guess that selection criteria vary a lot between users.

This might be a bit elitist, but, come on, MySQL? I heard they finally got transactions! Why not use Postgres instead?

- you’re right, my title was somewhat a bit of a teaser. But MySQL isn’t too far from achieving this level of performance actually.

- you’re also right about the developers running the database: but I guess the benchmark is thoroughly checked and balanced so the very same independent developers must have stressed the systems.

- Can you tell me more about RDBMS being better for certain applications than others? I didn’t know about that. If you feel like having an access on Tech IT Easy to blog about it, you’re more than welcome (long term thing man, no constraint whatsoever. Only rule = no rule).

- Wow, what an impressive list MySQL has. But you know, I don’t think you can just plug MySQL and see it work. You have to keep people working on it to support the load and the actual TCO of MySQL may be much higher than if proprietary databases like Sybase, Oracle, DB2 or MySQL had been used.

Interesting article, though I’m not convinced by sentences like “Microsoft SQL Server 2005 is twice as fast as Oracle’s Database 10g Enterprise Edition”.

Benchmarks might sometimes be a quite objective way to compare performances, but the performance of a given application working with a given DB depends much more from the developers than from the DB system! Furthermore some requests can be faster on a DB system, and other requests on another.

This has two consequences:

1 – a developer cannot make fast softwares if he doesn’t know its DB system really good

2 – a DB system can be great for an application, and unadapted for another.

Last but not least, I’m not sure there’s such a robusteness problem in MySQL. Just take a look here : http://www.mysql.com/customers/ There’s a lot of interesting links, which I think can be considered as solid references for MySQL

Xuoan> thanks for the links, interesting stuff from PlentyOfFish. No case study yet at my knowledge, but fortunately PlentyOfFish has many , many readers. I guess everybody knows MySQL isn’t scalable (although it’s a very nice database).

Remy> I should elaborate soon on the different modules included in SQL Server 2005. And you’re right: more and more big companies are migrating to SQL Server 2005.

When SQL Server was launched, a lot of people considered that it was dedicated to small businesses whereas Oracle and DB2 (IBM) were dedicated to big companies. With the recent events and the future version (SQL Server 2008), the RDBMS market will probably evolve on that point.

Plentyoffish is a nice example of SQL (and IIS) scalability vs Mysql :

http://plentyoffish.wordpress.com/2006/06/10/microsoft-aspnet-20-performance/

http://plentyoffish.wordpress.com/2006/10/08/digg-is-doomed-unless-they-fire-their-tech-staff/

http://plentyoffish.wordpress.com/2006/12/07/scaling-sql-server-2005-national-tv-usa-canada/

Maybe not a rule but a strong case study (maybe it has already been done @ Redmond)