About LWN.net

Notes from the Embedded Linux Consortium meeting

Inder Singh on ELC History and Directions

After seven months of existence, the ELC is coming of age. The ELC charter and board have been formalized, and the first election has been held. The ELC has been present at several trade shows, and has released numerous press releases. There are now 104 members of the ELC and were seven more signups at the Embedded Systems Conference.

A new market and technology convergence is happening with many new devices showing up. Many new processors are showing up, MIPS and Megabytes are becoming almost free. New System-on-Chip designs are showing up. The Internet phenomenon is happening and many connected devices are starting to appear. Connectivity is quickly becoming a de facto requirement for embedded devices. Software complexity is increasing. It is now common for an under-$100 device to require 100 programmers on the development team.

Linux is experiencing a previously unseen momentum, and embedded systems is where the action is. Linux is well placed to be the unifying platform for a currently very fragmented market. In this sense, Linux is in the position that DOS was in the 1980s.

There are daily software announcements in the mainstream press for Linux desktop applications. Observing Linux is the best means of tracking the technology curve. Linux is quickly becoming the operating system of choice for working with new semiconductors.

The post-PC era is emerging. Desktop systems are giving way to network connected devices. Linux has major traction at both ends of this spectrum. Cheap System-on-Chip devices, the Internet, and embedded Linux are enabling the development of network devices.

Predictions and recommendations:

• 2-5 years: more programs will be embedded than on desktops, Linux will be the predominant system.
• The ELC needs to move fast to insure the success of embedded Linux.
• ELC members need to find common ground and work together. Companies should be careful not to repeat the past mistakes of the Unix wars. Cooperation is vital and can be facilitated by the use of email lists and industry working groups. The embedded world needs modularity, hardware abstraction, promotion, and education. Communication and participation in the ELC is requested.

Rick Lehrbaum on the State of Embedded Linux

There was a poll and a count of hands, many more members' companies were actively working on embedded Linux projects compared to the previous year. A few ELC member companies from last year were mentioned. Many of these companies have gone from relative obscurity to being well-known names. Many of the small member companies have been acquired by big-name companies. To show the progress that has been made in the last year, Mr. Lehrbaum searched the net for historical articles on embedded Linux. Embedded Linux started showing up in news releases around June of 1999. These days, Embedded Linux articles are showing up everywhere. Linux Devices has been running several online polls, and Mr. Lehrbaum discussed the results. In short, Linux growth in the embedded field has been phenomenal.

John Cheuck on Linux in Japan

In Japan, Linux had a big showing in the LinuxWorld Tokyo conference. All of the large Japanese computer companies are working on Linux projects. Mr Cheuck commented that developers in Japan are having difficulties interpreting the OSS/GPL licenses. Despite this, a large growth in the Linux area is expected in Japan. Embedded devices are becoming more networked, this is a natural area for Linux. In Japan, there is a growing demand for skilled engineers with Linux experience, such people are already hard to find. The open-source nature of Linux makes it an excellent teaching platform, academia and industry can take advantage of this and focus on one system. In a survey of Japanese development engineers, these trends showed up: half of the systems being developed were 16- and 32-bit systems. Half of the programs being developed were above the 256KB target memory size. About 75% of the embedded programs were written in C, these trends tend to be very compatible with working in Linux.

Linux is well positioned to help with a much needed defragmentation of the Japanese embedded systems industry. Some large issues for developers of embedded Linux are where to find support and code maintenance, and how to achieve quality control. These are areas that vendors can enter and thrive in.

Emblix, the Japanese Linux Consortium was formed on July 13, 2000, the charters were finished September 26, 2000. The purpose of Emblix is to promote the use of Linux in Japan. Emblix will provide education and set standards. The chair of Emblix is from academia, Professor Nakajima of Wasada University. Emblix has three areas of interest: distribution, tools, and users. Currently Emblix has 61 members and 24 member companies.

Greg Wright on Non-Corporate ELC Membership

If you are involved in a small Linux company, this might be a very good time to get involved with the ELC. Many opportunities are to be had.

Murry Shohat on Moving ELC Forward

In order to achieve these lofty goals, the ELC needs to create momentum and promote embedded Linux in a vendor neutral and positive way. The ELC needs to continue to have a visible presence at trade shows. New members should be actively pursued, and there are plans to build an active database with helpful information for getting new member companies up to speed working with Linux.

Summary

The ELC would do well for its members by focusing energy towards clarifying the various licensing issues that developers face and translating the legalese into engineerese. Another area that still needs attention is the movement of embedded systems development tools onto the Linux desktop. Linux is currently showing up on many target systems, but the development systems are largely Windows based. A big gain in efficiency would be had by those companies who can provide and end-to-end Linux environment. Fortunately, a number of semiconductor manufacturers are already part of the ELC and can provide the expertise and motivation to bring about such tools.

The next few years should be very interesting as embedded Linux moves from development projects to working products and new successful business models involving Linux emerge.