Foreword: This interview with Yoshinobu Ushiyama, CEO of Lineo Solutions, is the third in LinuxDevices.com's CEO/CTO interview series. It addresses the growing momentum behind Linux in the Consumer Electronics (CE) market, and how the CE Linux Forum (CELF) can unify the market. Enjoy . . . !
Q1: How would you characterize the current state of embedded Linux, as a market and as a technology?
A1: Linux is gaining the position of the embedded operating system of choice among OEMs and has crossed the point where the tide will not reverse. The application of Linux to embedded devices has already passed the experimental and evaluation stages. The embedded engineers around the world are finding the way to overcome many of the shortcomings of Linux in embedded applications — especially in consumer electronics devices — such as larger than expected footprint, slow booting time, deterministic real time support, power management framework, memory management in critical sessions, refinement of code implementation method, robust file systems, and so on. Technologically, the activities to define CE Linux requirements by CE Linux Forum (CELF) should give certain direction and the framework for the development of embedded Linux.
Q2: What challenges do embedded Linux OS, tools, and services vendors face?
A2: The business model of the individual vendor is still under challenge and has yet to be proven. Vendors in the embedded market need to understand that the customers to satisfy are not the end users of the embedded devices, as in the case of PCs or servers, but the OEMs that manufacture the embedded device. For the end users of the embedded device, what system runs the device has no significance. A high level of embedded device development know how needs to be delivered to OEMs in the form of tools and services. When I consider the variety of embedded devices and the variety of CPU cores for embedded devices, without providing engineering expertise in some form, an elaborate universal tools product might not be sufficient to increase productivity at the actual embedded product development site of OEMs.
Q3: What market opportunities do you see for Linux in the embedded devices and systems market?
A3: The GPL and Open Source nature of Linux should free OEMs from dependency on a specific vendor. This will also provide an open opportunity for chip vendors to focus their technologies and expertise to meet specific embedded device market needs with Linux. The embedded market covers the full spectrum of industry, with a wide range of devices and systems. So long as the socio-economic tide toward the Ubiquitous Network Age prevails, the opportunity of Linux in the embedded market will be open for anyone to realize.
Q4: What challenges does embedded Linux face as a technology?
A4: Considering the amount and quality of human resources involved in the communities and in the embedded industries, embedded Linux will be able to answer any technological challenges. Therefore, actual technological challenges boil down to how to contain the deviations in the implementation of a solution within the framework of Linux and Embedded Linux. Some may call it the fragmentation issue; others may call it the Good Fork-Bad Fork issue. In the spirit of Open Standard/Open Source nature of Linux, I am, however, against any move to discourage the innovative motivation of the developers. The issue is how major technologies around Linux, especially embedded Linux, can be accepted as universal standards by the industry. The effort by CE Linux Forum working together with Open Source community could provide a practical framework so that we don't have to see again what we have seen in Unix.
Q5: What embedded Linux technology developments do you find exciting?
A5: When I limit myself to embedded Linux, which covers such a wide range of applications beyond PCs and servers, any technology to expand the scope of battery-operated devices, regardless of the field of application, is the most exiting. Deterministic real-time support by embedded Linux for industrial applications is also very exciting, because Linux will be able to provide the standard network connectivity for all industrial devices and make industrial systems an important part of infrastructure for the Ubiquitous Network Age.
The announcement of Consumer Electronics Linux requirements by the CE Linux Forum, with some code implementation efforts made by CELF, will open new opportunities for a realistic approach to the solution, especially when the embedded Linux code base will move to the 2.6 kernel.
Q6: Can you share one or two of your company's most exciting successes?
A6: Lineo Solutions, Inc. was born only a year ago, so I don't have much in the record in the name of Lineo Solutions. However, we engineered hand in hand with Sharp to put the SL Zaurus in the market. This engineering effort was not limited to Linux kernel porting, but extended into the actual integration of PIM and JVM software. I think, this effort by Sharp and our engineers proved to the Linux community and embedded industry alike that Linux really can be used in embedded applications.
I consider the contribution to CE Linux Forum of x86 embedded Linux code by Lineo Solutions to be another achievement. This will make it possible for anyone to evaluate the CELF embedded Linux code implementation against Linus's Standard Linux on inexpensive x86 reference platform hardware easy to obtain anywhere in the world.
Q7: How about a failure?
A7: Lineo Solutions is in the business for just a year as a newborn Lineo. Therefore, there is no case of failure, yet.
Q8: What's your vision of the future for embedded Linux? That is, how big will the embedded Linux market get, and when will it start to reach that peak?
A8: Embedded Linux has just gained the initial and serious momentum at last, to my eye from the third quarter of 2003. The embedded Linux is now here to stay as the operating system of choice for embedded devices. When you look at the scale of CES in Las Vegas, it already exceeds the peak of COMDEX some years ago. Honestly speaking, how big the embedded Linux market will grow is for anybody to guess, I think.
As to the peak, it all depends on how deep and far the social change will be brought about by Ubiquitous Network Age, and since the door to the Ubiquitous Network Age has also just opened, I think it is beyond anybody's imagination at this stage.
Q9: How do you see the Eclipse tools platform fitting into the future of embedded Linux development? Is Lineo's “Elite” cross-platform Java development tool based on Eclipse?
A9: No, ELITE is not based on Eclipse. Even though ELITE IDE is developed as Java application, ELITE is not a Java development tool. Eclipse provides a universal framework for all the application development needs for IT Industry as a whole. I, however, have a notion that IT Industry is an assortment of many industries that cover the whole spectrum of socio-economic activities. Today, IT industry means business, any business. Of course, I see the need for and great merit of having the general framework — you may call it environment framework — for development tools of business applications and services that will be built on the ubiquitous network infrastructures. But device manufacturers in the embedded device industry need dedicated tools to increase productivity in designing and manufacturing devices. ELITE is aiming at satisfying those needs. When everyone is basing the tools on Eclipse, I believe there is a need for dedicated and focused tool to embedded device development.
Q10: How do you see the 2.6 kernel affecting the embedded Linux market?
A10: 2.6 kernel opens up a new technological front and business opportunity for embedded Linux. Incorporation of pre-emption and uClinux expands the adaptability of Linux for wider range of embedded devices. This new opportunity will reinforce the dominance of embedded Linux in the embedded industry.
Q11: Any comments on the SCO mess?
A11: Much of the concern with this subject is in the media and among market analysts, I think. Once upon a time, not so long ago, there was a champion of Linux and the Open Source Movement. It lost the DNA of its inception during its growth, and became mutant. It's a pity. But my comment here is only a sentimental one, and nothing more.
Q12: What do you think about Wind River's recent public endorsement and announcement of support for embedded Linux, and recent decision to join the Consumer Electronics Linux Forum (CELF)?
A12: Since the foundation of CE Linux Forum in last July, embedded Linux gained unstoppable momentum to be the choice embedded operating systems for the Ubiquitous Network Age. I had a respect for Wind River's technology and solutions because we were embedded system developers before we come into the embedded Linux business. The move by one of the major detractors of Open Source and Linux to Linux Camp is good for the embedded Linux as a whole. How they are going to overcome their history as the champion of proprietary embedded OS and tools, however, remains to be seen.
Q13 Lineo Solutions joined the Consumer Electronics Linux Forum (CELF) last September, and appears to be a very active member. What are your thoughts about CELF and its mission and goals?
A13: CE Linux Forum was initiated by concerted effort of leading OEMs, users of embedded Linux. I think that is significant. Embedded system and software vendors, and chip vendors, they all have customers. In other words, embedded system and software vendors and chip vendors have to satisfy the needs and requirements of the customers. Embedded Linux is not the exception to this. The complexity of embedded systems, software and chips alike, is rapidly increasing so that no single vendor is self-sufficient to provide the total solution single handedly, especially in the GPL Linux world. Lineo believes that CELF CE Linux requirements currently announced and going to be announced in the future from time to time will give certain direction to focus our development efforts as an embedded Linux vendor. Since Lineo has been committed to OEMs to put embedded Linux devices everywhere in the market, Lineo was prepared to make contributions to CELF even at the early state of its activity. And Lineo will continue to make further corporate and engineering efforts of contribution, and do our share as an embedded Linux vendor, in the future.
Q14: Lineo Solutions recently joined the ARM Connected Community. Can you tell us more about that organization, and Lineo's role within it?
A14: ARM connected community is a consortium of third party vendors to support ARM licensees and ARM users, therefore it is not limited to Linux vendors. The ARM core has a strong customer base in consumer electronics devices, and strong Linux support is expected. As you know, most ARM core chips are SoCs, and this trend will accelerate. Therefore, it is important for Lineo to properly support frameworks for memory management, power management, etc, provided with the architecture, so that the device developer can utilize the best of the chip's performance. These are part of the effort by embedded Linux vendors around the ARM connected community.
Q15: Can you tell us about any customers and projects you are helping them with?
A15: There are many projects underway with OEMs, but all of them are under NDA. We cannot tell what the right hand knows to the left hand.
Q16: How is Lineo Solutions's transition away from Lineo/Embedix going?
A16: Actually, we have completely transformed to a newborn Lineo already. Now, Lineo Solutions is re-established as an embedded Linux distributor, with Lineo uLinux that supports most major embedded chip architectures. We also have an IDE product of our own named uLinux ELITE. Lineo Solutions put over 18 years of experiences in embedded business into uLinux and ELITE. For a Linux vendor, 18-years experience in developing embedded devices is rare to find. This was possible because Lineo Solutions did not loose a single engineer at the separation from Embedix and during transition period to become Lineo Solutions. Lineo Solutions did not have any negative technological and corporate consequences around becoming an embedded Linux vendor for the Ubiquitous Network Age. That was one of the reasons we could directly contribute portions of the Linux code based on Lineo uLinux to the CELF Source Tree even at the earliest stages of CE Linux Forum activity. Lineo Solutions intends to keep contributing a lot more to CELF so that OEMs can accelerate putting embedded Linux devices into market.
About the interviewee: Graduated from Nishi High School in Tokyo, Ushiyama was enrolled in Northeast Missouri State University in the United States, then spent two and a-half years in Eastern/Western Europe. As he came back to Japan, Ushiyama engaged in folkcraft wooden furniture manufacturing in Matsumoto, Nagano. A few years after, he moved to Tokyo and worked for Paris Miki Inc., a major Japanese retailer of spectacles (glasses), and then started his own retail glasses business. When he came back to Matsumoto again in 1984, Ushiyama co-founded United System Engineers, Inc. (USE), along with Tatsuya Takeuchi.
Brief history of Lineo Solutions: Yoshinobu Ushiyama co-founded United System Engineers, Inc. (USE), along with Tatsuya Takeuchi, in Oct., 1984. In May of 1999, USE was acquired by Lineo, Inc. and became a wholly owned subsidiary of Lineo. In July of 1999, USE was renamed as Lineo Japan. In Apr., 2002, Embedix, Inc. reincorporated Lineo, and Lineo Japan became a subsidiary of Embedix. In Dec. 2002, Ushiyama and the employees did a management buy-out of Lineo Japan from Embedix. In Feb., 2003, Lineo Japan changed its corporate name to Lineo Solutions.
This article was originally published on LinuxDevices and has been donated to the open source community by QuinStreet Inc. Please visit LinuxToday.com for up-to-date news and articles about Linux and open source.