Month: April 2009

Stunning news today that Oracle has offered to buy Sun Microsystems. This is sending the MySQL community reeling, as they begin their MySQL Conference & Expo today. Everyone’s talking about whether this change is good for MySQL.

My first thought is: it’s not over till it’s over. These deals have a way of falling through at the last minute. Just look at Microsoft’s attempts to acquire Yahoo!. I’m not saying that it will fall through in this case. Just don’t count it as a done deal until the agreements are signed, and the shareholders and the SEC have their say.

My second thought is: it depends on how well the two companies can integrate. In any acquisition, there’s a merger not only of assets but of goals, strategies, and corporate culture. Not to mention people. People matter.

I worked for InterBase during part of the 1990’s. InterBase was an RDBMS that was acquired by Borland in 1991, as part of their acquisition of Ashton-Tate. Borland was very interested in Ashton-Tate for its dBase product, but Borland also got InterBase in the deal (InterBase had been acquired only a couple of months before AT’s merger with Borland). InterBase wasn’t in Borland’s strategy and it wasn’t what they valued as part of the acquisition. As a result, it was an unwanted step-child for over ten years (despite having a revenue matching Borland C++Builder).

What does this tell us about Oracle’s plans for MySQL? Nothing for certain. My point is that it depends on what Oracle values as part of the acquisition. Is it Java? Is it the line of enterprise hardware? Is it XFS or OpenSolaris or NetBeans or Glassfish? Any of these are likely candidates. But MySQL does not jump to the head of the list as the likeliest “jewel in the crown” that motivated Oracle to make this offer.

In his blog, Vadim reports that a new storage pluggable engine for MySQL has appeared in the source tree, to support IBM DB2 for i as a back-end.

This reminds me that I hate the IBM System i platform (aka IBM Power Systems, aka iSeries, aka AS/400).

Don’t get me wrong — I’m sure it’s terrific technology. I’m sure IBM supports many businesses with it and they’re happy (although I do wonder why they need to keep re-branding the product line). But my fate is not aligned favorably with respect to System i.

At several companies I have worked for, the business development people struck an ill-conceived deal with IBM, to “support System i.” Meetings were had. Agreements were signed. Commissions were paid.

Then it came time to do the work and fulfill the partner agreement. At my last job, my manager came to me and said, “by the way, in your spare time, make sure your work-in-progress supports the IBM platform.”

I know nothing about the IBM System i. I have never seen one face-to-face. I have never seen any documentation for it. I enjoy command-line interfaces, but using the System i made MPE/ix seem friendly.

By the way, in spite of the “Universal” moniker, DB2 on the System i is, as far as I can tell, a completely different database implementation, with the brand name “DB2” tacked on as an afterthought.

Here are some suggestion for the System i business development folks at IBM: when you make a deal with small companies to support your platform, make sure they have enough machines to do development and testing. Include electronic documentation so everyone can have access to it. Perhaps even offer some training as part of the deal. And then ask your new partner for a project plan that details such things as:

• Which products they promise to support on the System i.
• When they promise to do the work and have the solution ready.
• Who they will assign to do the work, not ask to do it in their spare time.