development

Data Center Knowledge has a story http://www.datacenterknowledge.com/archives/2009/11/25/wows-back-end-10-... about the current World of Warcraft infrastructure

That is just awesome. The bottom of that story has additional links to information about Second Life and Everquest that are interesting too if you are into that.

If you have no idea what this is let me describe some things.

In the Massively multiplayer worlds you rarely have a single instance of a world with all of your users in it. So in this case what you will have is hundreds of world instances in clusters and you and your friends join a given instance to play together.

Imagine the game world like like this:

main game &
login servers ---> instance1 -------> server1/server2/db1/db2
                    |
                    |> instance2 -------> server1/server2/db1/db2
                    |
                    |> instanceN -------> server1/server2/db1/db2

And so on. Plus of course dozens and dozens of other servers for all of the non-world infrastructure.

At say 30 watts per CPU core assuming multicore processors and say at .10 kWh they are spending about $2 million on power alone. I am glad the players are sitting alone in dark basements, maybe it equals out on the energy production side :)

This article covering a presentation on how to grow your site covers several points that I hear over and over again when talking to folks about building products.

http://carsonified.com/blog/web-apps/9-ways-to-take-your-site-from-one-t...

The ones that stick out of every talk and presentation are the following:

1. Pick your core features and get them out

2. But don't think you know exactly how the users will use them or that you have them right.

3. Learn, iterate, learn, iterate.

4. Test, try, build something, see if it works for you and your users. You will not be all knowing.

There are a few more things that I want to add from my experiences:

5. Your customer often doesn't have any more idea than you, so don't think he isn't just spitting out random bits.

6. If you have a new product or are trying a new space, do believe in your idea and don't get dragged all over by folks trying to make your product in their image.

7. Idea's are great, but they take time to try. Focus on how to get a win before you try every adjacent market.

8. Make sure you actually know how to find your customer. They have no idea about your better mousetrap until you lay out some cheese.

9. Do you have any idea how to convert a customer, what are they buying? why they are buying?

10. Your customers or your company will get some bad apples eventually, don't worry about your awesome structure yet. Just get it going.

As illustrated here, http://www.calnewport.com/blog/?p=115 , in a post about The Einstein Principle, "We are most productive when we focus on a very small number of projects on which we can devote a large amount of attention."

Now stop reading blogs and go focus on something.