IBM Exec Sees Open-Source Interest Growing

Bob Sutor, IBM's vice president of standards and open-source, sees 2006 as the year when a number of industries will move to embrace open-source software -- and he expects IBM to play a role in many of those efforts. Sutor, who spoke with Computerworld Monday, also weighed in on the role of Linux and offered his thoughts on Massachusetts' OpenDocument plans.

What markets will embrace open source this year?

I think there will be a continuation in health care and education. Those industries have tremendous social relevance and also have significant amount of work to do. We will see a lot of government policy related to that.

We're also starting to see more open-source projects that are very specific to a particular industry. The SAKAI project in education is a good example. It is basically an ERP system for universities. Instead of managing customers, it manages students and courses and allows information sharing between colleges and universities. IBM has been advising them.

I do think financial services will have a big year. With any technology, financial services gets in early. They're always looking for efficiencies and greater economies of scale. Retail is another area, as we see a lot of standardization work going on there. These are the hot areas we're looking at.

Speaking of retail, Windows still holds more than two-thirds of the installed base of point-of-service terminals in stores, with your 4690 systems still holding a big slice. Linux hasn't made much progress despite its strengths for a thin client -- which is what a POS terminal is. On the Linux server side, we've done some interesting things in stores. There's an RFID project with German retailer Metro AG and SAP. Linux servers monitor what's happening on the shelves, like when items are taken off and brought to the register. While I certainly believe we'll see more Linux infiltration in terms of the standard things people are doing now, I think as retailers get more creative and start building these "stores of the future," they are going to look for Linux. The price is right, the ability to code on them is right.

The world is going to be hybrid for a long time. People need to view it as an optimization problem: What is the right mix of Linux, other open-source and proprietary [software] in terms of the economics, my staff, my partners? Anyway, I wouldn't ask, 'How do we get more Linux in retail?' It's how do we get it thoroughly open-source and standards-based so there's a level playing field. That way you can see where open-source makes sense, and where proprietary products make sense.

Do you still hold out hope that we'll someday see an open-source Java?

You may remember we made a big to-do about this two years ago, with an open letter to Sun. I think the way things are going on Apache, we're getting there little by little. We said what we said two years ago because we wanted to get the issues out there and have Sun respond. In different ways, they have responded.

I think people have to realize that certain programming languages are optimized for certain activities. There are still billions of lines of Cobol out there. There's a lot going on with PHP. Ruby is coming along. There are new programming methods such as Ajax. The positive thing is that there looks like there's a lot of innovation going on in the programming language and tools space.

IBM just submitted something around Ajax for Eclipse. Eclipse is written in Java. We didn't have to think politically, 'Oh, Eclipse is written in Java so....' It's the right programming language for the right job. But as long as innovation continues, and the tools become better, that's fine. Things will shift naturally, wherever they need to go.

So it's not a matter of Java having not fulfilled its promise because it hasn't gone open-source -- it's that some of these languages are technically more optimized for certain applications?

That's right.

What's your prognosis for Linux on the client, especially now that virtualization technologies may give it a boost?

The Open Source Development Lab late last year had a meeting to really start to figure out how to converge some of the technologies on the client. Progress is being made, and OSDL has really taken a leadership position on this.

You must be pretty disappointed now that Massachusetts looks like it may take a watered-down version of OpenXML. You've read some headlines that are incorrect. The state is firmly committed to moving forward with its OpenDocument plans on Jan. 1, 2007. Microsoft made some moves at the end of last year. The state said basically, 'We'll look at them as they come along.' As the situation changes, if Microsoft commits to doing everything that they need to -- and I would argue that they have not done everything that they need to do -- then the state will look at what they do like anything else. But in terms of what they've said about OpenDoc, the state has renewed its support for that and that's the assumption we're making right now.

But Microsoft seems like it has made some inroads. The playing field seemed to be that you and others were trying to cement OpenDoc as the primary and only format, and now there may be more than one. The reference model that was put in place last September and is still in effect is only OpenDoc. Microsoft has given their formats to ECMA and basically told ECMA, "Here they are. Don't change them, but please call them 'community and open' even though we're the only ones allowed to do anything with them." They're hopeful that this will somehow get a nice rubberstamp that will convince people to do something with them. And all Massachusetts has said is, 'Yes, you've done something, we'll look at it next time comes around to see if it meets our requirements. If it doesn't meet our requirements, well, then it's still out.'

If Microsoft does make it sufficiently open -- and I would argue it is not -- then Massachusetts will do something. But it's in the future. Don't fall into the trap that something that may happen in the future means that OpenDoc is out. There's a lot of work to be done in terms of Microsoft formats, ECMA formats, before they become a standard.

I think there's going to be a lot more news this year around OpenDoc adoptions. The struggle is not over. You have to admire the pioneering efforts of Massachusetts. But when we look back, it will be the first of many.

Is IBM pushing OpenDoc hard in Europe or Asia?

The whole globe. What was established in Massachusetts was an excellent amount of collateral: Why, from a technological and open standards perspective, OpenDoc was a superior standard. There was a tremendous amount of FUD generated around OpenDoc, all of which was dispatched fairly quickly and with great precision. What's happening in Massachusetts will make it easier to go to other governments, because fundamentally the arguments are quite sound, and have been developed over months.