No this post is not about Sun buying mySql.

So I compiled a fresh 5.0.51 on two boxes. It’s quite easy to setup, all it takes is the main server and the sub servers to support federated engine.Only trick is to remove fulltext indexes on the federated tables. Leave the main server as-is. On the sub server, create the table with ENGINE=Federated and set the connection to the main server. Make sure you have the user created on tha main server and the password of the user does not contain any special characters. (some reported @ $ characters are problematic).

It works on the first try. Great feature! However, setup is a bit long if you want to federate a whole database with 100 tables in it.

Then you try your first query on a mid sized table with 250K rows.

SELECT * FROM table_t LIMIT 10;
Runs 0.03 sec on main server, takes 8 seconds on the federated server!

then the second:

SELECT count(id) FROM table_t
Runs 0.02 sec on main server, takes 252 seconds on the federated server!

Looks like I’m not the only one complainig about the problem so I didn’t care to fill in a bug report.

More info about federation is here if you are still interested in.

Update:
Giuseppe Maxia contacted me refering his article on federation engine. It’s good read.

Congrats to Aral Balkan and his team.

The irony is that we argue about whether systems should be application-centered or document-centered, probably the two most important entities in a computer. Have you ever seen a system which lets you, out-of-the-box, hit a corner in order to do anything at all even remotely related to anything having anything at all to do with a document or application? So maybe documents aren’t the most important entity in a computer. Browse the internet by hitting the screen corner? Check mail in the screen corner? Get Info in the screen corner? System preferences in the screen corner? Switching applications in the screen corner?

And this argument makes more sense after seeing SymphonyOS: a Desktop computer operating system based on Debian GNU/Linux and Knoppix GNU/Linux.

I seems some people out there is really working on human computer interfaces…
Great job. I’m impressed.

Some titles:

A buyer’s guide to open source
Open source business intelligence
Open source business process management
Open source content management
Open source CRM
Open source ERP
Open source enterprise service bus
Open source identity
Open source directory
Open source portals
Open source point of sale
Open source RFID
Open source VoIP/Telephony
Open source licensing offers many choices

Here is the link

Days passed and I see the things much clearer.

After reading this passage, you’ll understand the Microsoft way and Microsoft’s idea about computers and computer users (plus developers). The question you should ask to yourself is “What is an Operating System?”
According to Encyclopedia Britannica Operating system is :

Software that controls the operation of a computer, directs the input and output of data, keeps track of files, and controls the processing of computer programs.

With one sentence maybe this is the best explanation you can find.
My point is, an operating system’s duty is not supporting custom operations. For example, your can not expect an operating system to have an internal spread sheet support or word processing support.
Those are called: Applications or simply software and from Britannica: Instructions that tell a computer what to do.

Quote from MS RSS developers’ manager: “We want developers to build good applications rapidly without going all through the hassle of rewriting lots of code.” (I don’t really remember what he exactly said but his point was this.)

This is a great objective for a project! However, is operating system level suitable for this kind of support?

Generally when developers write their codes, if they plan to re-use the same code again and again, they put the re-usable code together and call it a library. They share these libraries with each other and call it open source. Most of the time the reason they share is just for the sake of sharing or feeling sorry for the future developers who will go through the same problem. This is not a foreign term for most of us. For an open-source developer it’s the foundation of all software. (Java, C, PHP, Perl, Python and many more all live with shared libraries.)

However MS mind has a different point of view. As it seems they see the open-source community not useful way for software evolution. Instead of allowing people to access open libraries they hide their code. They bury their code deep under the OS to hide it.

It’s the main reason that any MS operating system will not be better, faster and useful then any simple open-source operating system supported with vast open-source libraries. When you boot your MS box, it will deal with huge amount of garbage that you’ll never even use through all your life. The OS will try to handle hundreds of thousands (soon possibly millions) if API calls and not give you a hint of performance and usability.

When you need a support for a special application for like RSS you don’t build it on OS level, you “extend” the OS with a library. But it is so far for MS to build a library and distribute it for free. They want us to pay for every bit of the code they wrote. That is MS point of view.

In the end I can not blame them not to give anything for free. They are a company for making profit. But I can’t let anyone to tell me that :
“Longhorn loves RSS”.
No one is yet stupid to believe this.What I’d say :“Longhorn loves money”.