Interview -- Hiroki Kamata and Junichi Suzuki Representatives of OMG Japan

--interviewed by John Sachen and Dirk Slama

In a conversation with the authors, OMG's Hiroki Kamata and Junichi Suzuki give some interesting insights on the situation of the CORBA market in Japan.

John Sachen: Kamata-san, you are the OMG representative in Japan. From OMG's perspective, what do you see as the main differences between the middleware market in Japan and other countries like the USA?

Hiroki Kamata: In one way, the situation is very similar. After the downsizing hype of the early 90's, companies have sobered up. They have realised that client/server is not a simple technology, and that their so-called legacy systems are indeed IT assets; those systems that are still running today's business. The keyword of the late 90's is "enabling diversity:" most end-customers are locked into mainframe technology provided by hardware vendors like IBM, Fujitsu, and Hitachi, and they are under extreme pressure to leverage and integrate these systems with the rest of their technology park.

The hardware vendors, on the other hand, have discovered that services and technology integration is an extremely lucrative market in a post-Microsoft world. IBM's first quarter results in 1999 have shown that IBM is no longer a machines company, but has successfully transformed itself into a services company. The Japanese hardware vendors are clearly following this trend.

Vendors like Fujitsu and Hitachi have adopted OMG's CORBA as a key platform for technology integration. And this is where one big difference between the USA and Japan actually comes in -- while business in the USA is always highly competitive, we can actually see different vendors co-operating in Japan. They are very actively promoting cross-vendor interoperability, to prove to customers that CORBA technology actually delivers what it promises.

Sachen: Who are the main players in the Japanese CORBA market?

Kamata: In the beginning, the Japanese CORBA market was dominated by western players. IONA Technologies had a very strong headstart in Japan. Two years ago, Fujitsu was sceptical about CORBA. Now Fujitsu and Hitachi are rapidly gaining ground. They have each assigned over 1,000 people to CORBA-based system development. Hitachi plans to base 60% of their SI business on CORBA and component based middleware by 2002. Other key players now include UNISYS and OGIS (Osaka Gas Research Institute).

Dirk Slama: Does the demand for technology integration seem to be a main driving force behind the success of CORBA?

Kamata: Globalization and the Internet seem to be the two most important events this decade. CORBA as an Internet technology is well positioned to support Electronic Data Exchange, e-commerce, and business-to-business transactions. Mergers and global alliances require companies to integrate highly diverse IT infrastructures. CORBA was designed from the beginning to deal with heterogeneous, distributed systems.

Slama: Where do you see the bottleneck for the growths of CORBA technology in Japan?

Kamata: This is clearly a staff issue. I think the biggest problem is a lack of experienced system architects, designers, and project managers, who are able to cope with the technology mix. Many CORBA projects combine mainframes, Unix, Windows, Java, C++, COBOL, relational databases, etc.

Slama: Is CORBA too difficult?

Kamata: No. CORBA itself is a relatively simple technology, easy to pick up for programmers. Since CORBA is component-based, individual programmers can focus on small subsets of complex distributed systems, so this is not the problem either. The problem is the lack of "big picture" people who can deal with the technology mix as a whole, who are able to design the component architecture, and manage the different development teams.

A lot of this has to do with hierarchies in Japanese corporations, education, and the typical Japanese career path: first you do ten years of coding, then you spend 10 years in system design, and finally you get to be a senior consultant or system architect -- using 20 year old technology?

Junichi Suzuki: In the age of the Internet, you have to take a risk to be on the cutting edge, but this is not the Japanese way of doing things.

Slama: In the US, CORBA has become a well-established mainstream technology. What is the impact of the technology gap between the US and Japan?

Suzuki: If I look at CORBA implementations, I would say there is literally no gap. Vendors like Fujitsu, UNISYS, and NEC are offering very robust, scaleable CORBA implementations, made in Japan. The problem comes in at the end-user level, as discussed before -- lack of experienced architects and project leaders who are capable of leveraging the strengths of CORBA.

Back to Main Article - CORBA - The IT Garage

Comments or suggestions?
Contact cjmaster@cjmag.co.jp