Keith Watanabe * NET 2.0

It wasn't that big of a surprise to me when Oracle purchased BEA.  It's been said that BEA has been looking for a buyer for sometime now as Open Source has challenged their position on the application server market.  In some ways, this is a natural pairing, considering that a lot of the companies who use BEA also use Oracle (because they have the money).  It'll certainly be scary in some ways for companies that employ both solutions since Oracle moves in to slowly control the application server market, but at the same time I figure that those companies probably can afford that type of solution.

However, I think Sun buying out Mysql is a horrible move.  It's a good finacncial move for Sun, considering that Mysql probably is the most popular RDBMS out there.  However, in the past 10 years, I haven't been all that impressed by Sun's track record.  I'm certain that anyone who's read this blog would note that I have a particular thing against Sun's so-called vision (which is just marketing-speak for throwing things against the wall and seeing if they stick).  And I've been extremely vocal against Jonathon Schwartz as Sun's latest CEO.  Right now, he's been credited as someone who has resurrected Sun after McNealy's departure and that this move was all due to him.  That said, I think they're going to bungle Mysql and it's going to hurt the Open Source community in the long term.

First, doesn't Sun already have Ingress?  Why not support that?  Does Sun's move in buying Mysql imply that they'll let Mysql rot?  Will they continue to commit to allowing Mysql to be Open Source and free for most cases?  Or will they simply start enforcing new licensing policies that have the potential for decimating a good portion of the web?

I'm sure that many other Open Source developers and companies who rely on Mysql probably have similar contemplations with regards to this move.  Or perhaps a better word is fear. 

The interesting part about the move was that Oracle's move came the same day.  Coincidence perhaps?  Sun's goal with J2EE was to always attempt to move customers away from RDBMS and onto hardware (their hardware!) solutions by creating the notion of a middleware platform.  Oracle's move with BEA shows me that Oracle wants to completely dominate the enterprise market and high and mid level tiers.  With Mysql in Sun's court, it's clearly a threat to Oracle's position on the business application market.  Sun could theoretically move to completely supporting a Java Open Source version of the application stack with Mysql + Tomcat + Hibernate.  If they continue the current licensing with that stack, Oracle will find themselves in an even more limited situation, considering they are one of the most expensive software companies around.

From there, we have to look at the rest of the market for this.  Obviously, we have LAMP which just isn't owned by anyone.  That's the purest form of Open Source available and probably the most dominant.  You have Microsoft's .NET + SQL Server bundle.  Then you have Redhat + JBoss/J2EE + LAMP and support for Postres.  IBM with Websphere + DB2 and major support for Suse and Redhat.  Then Sun with Sun hardware, Mysql + massive Java support.  And now Oracle with Oracle + WebLogic.

Some part of me thinks that Sun's purchase of Mysql is for their grid computing project.  Maybe have a massively distributed, rental system like N-1 that contains an application stack that they support.  So companies will never have to worry about support again in terms of reliable system administrators, DBAs, and production support.  Instead, companies will just need to focus on development utilizing a well priced fee for their services.  In other words, a highly trustworthy hosting service focused on providing Java solutions.  They probably could integrate many enterprise features like their user management tool Liberty.

If Sun takes this route to computing, then it would actually not be a bad idea.  I've proposed that Yahoo (and Google to some extent) should provide similar services, in evaluating their level of expertise on web scalability matters.  In Sun's case, the cost for them would be low since they don't have to worry about purchasing hardware nor licenses as everything is internal.  In some ways, I would consider this move to be something like enterprise hosting.  In this vision, large firms could rent out space from Sun's data center where the entire application stack is provided.  Sun would provide all the support and the firms would migrate their applications onto such a platform with Sun's help.  In house staff could be eliminated or moved into Sun as an agreement.

Anyway, if this is the vision for Sun, I think it would be a good business move overall.  It depends on whether or not they're wiling to focus on it and take such a huge gamble that companies would want a service like this.  Also, it would be scary in some ways.