As the chief open-source officer at Sun Microsystems, Simon Phipps spoke to ZDNet Australia about the MySQL acquisition and community engagement on OpenOffice.org and OpenSolaris.
Q: In the beginning of 2008, Sun spent $1 billion on the
Phipps: It's a bit soon to be making that sort of judgment. Asking that question now is a bit like asking a company to change its product strategy on the basis of the share price.
MySQL is a long-term acquisition that has to do with the restructuring we have done of the company. It has to do with being part of a holistic solution, if you look at what we have done at Sun, in November we restructured the company. We have restructured the company so that now open source is at the heart of the entire product business.
We have three divisions. (Our first division is) an applications platforms groups. We have Java E, Java ME, and Database with MySQL in it, their business model is to drive adoption and then monetize subscriptions.
We have a second division, our systems division, and you could say their business model is to use open source as firmware.
The third business is using open source as the backbone for cloud computing, so we have a cloud computing division that uses open source and then monetizes the delivery of both software as a service and software as a service infrastructure.
The MySQL acquisition fits into that, both as a product set, and as access to the install base of MySQL and also for the open-source skills of the executives and staff that we hired.
I think now is too soon to be doing a basic mathematical judgment. I'd say that from an organizational point of view it was an ideal acquisition; it was one of the best acquisitions that Sun has done in recent years.
But is it safe to say that the MySQL acquisition hasn't been profitable so far?
Phipps: No, I would say that is a very unsafe thing to say. I would say that .
Moving on, Sun has just released
Phipps: There has been 35 million downloads since October. When you look at the historic graphs showing downloads that is significantly up even on a new release.
Anyone who is assessing the success of OpenOffice.org 3 would be making a huge mistake if they fail to account for the plug-ins, where there has been more than 50 new plug-ins since October.
Given the size and complexity of the OpenOffice code base, have you had any luck recruiting developers outside Sun for the core code?
Phipps: The complexity of the core code base is such that it is a specialized task to engage in modifying the core code, but the big innovation in OpenOffice.org 3.0 was the introduction of the add-on mechanism. There has been a significant number of developers who have engaged in that add-on mechanism.
Even more importantly, we are seeing large enterprises adopting OpenOffice.org 3.0 and creating add-on for their own applications.
I would be the first to say, given my role is community engagement, that the code around OpenOffice.org is hard to approach. That's why we have implemented this new plug-in mechanism.
So do you have developers outside Sun working on the core code?
Phipps: There are developers working on that core code, notably the developers from IBM, who are using the code for Lotus Symphony. They have some catching up to do.
They, and also a number of developers from Red Flag 2000 who are working on China's Red Office, are all engaged in a community working on the code.
If you add up all the individual developers who work on the OpenOffice code because they have consulting engagement or small businesses, that sector is almost as numerous as the Sun developers who are engaged.
The second highly significant area is localizers. There is a very large localization community that is doing local versions of OpenOffice. I believe there is OpenOffice in over 100 local languages around the world.
One of the great values of OpenOffice that is really remarked on in analysis is that OpenOffice is localized globally. That makes it a very powerful force in the industry because it is able to go places with community engagement that proprietary software can't go.
Phipps: Yes, we have had a steady trickle. The obstacle on OpenSolaris is not so much the complexity of code.
The problem with OpenOffice is that there is a very strong dependency between every part of OpenOffice. That means that when you make code contributions, there are very few developers with the oversight to be able to accept those code contributions, and that causes a bottleneck.
In OpenSolaris that is not the case, OpenSolaris is a very well structured and diffuse piece of code, where it is very reasonable to be able to have contributors.
Alex Serpo of ZDNet Australia reported from Sydney.