At the Start of TRON Forum
Ken Sakamura (TRON Forum Chair)
I developed the kernel of real-time OS under the name of “The Real-time Operating System Nucleus" since the beginning of the 1980's. Initially, the specification was released based on the philosophy of open architecture. Then even the source code was released after a while.
This can be called the early software in the beginning of "open source" era.
When it comes to "open source", simply releasing the software as is after you develop it is NOT enough.
First of all, can people other than the original developers understand the source code?
It is a problem of readability in a nutshell.
To improve readability, we need documents such as external specification of the source code and internal specifications that describe how the code is structured.
Suppose we solve the documentation issues. Then comes the distribution hurdle.
When the project started, the Internet was only available to select academic institutions, and it was an age when the personal computer communication had become popular among dedicated hobbyists.
In the computer industry, the specifications were printed on paper and then distributed.
A contact address manned by people was necessary so that the bug reports or suggestion for improvements could be accepted.
If you wanted a wide acceptance, participations in symposiums, and offering training seminars became necessary, too.
It was an era when you needed more time and man-power in contrast to today's situation if one wanted to release open source software and wanted it to succeed.
One needed money for this, too.
At least, I noticed the need for an organization to support the activities when I started TRON Project in 1984.
Clerical man-power was necessary on top of engineers.
Naturally, we needed an office space, too.
However, an organization to release its project result free of charge is naturally a not-for-profit organization. So we could not invite investors to start its business.
To make a long story short, Japan Electronic Industry Development Association (JEIDA merged into Japan Electronics and Information Technology Industries Association in 2000) stepped in and was kind enough to offer support staff voluntarily for the project activity. This was in1986.
This was the start of informal TRON association.
However, JEIDA could not be counted on forever: after all, it is the organization for the promotion of electronic industry of Japan although it had a focus on the computers.
We needed to create an organization of our own, and so TRON Association was established in 1988.
At that time, there was no concept similar to today's cloud funding although we would have welcomed everybody who chimed in with their share of contribution.
Back then in Japan, we had to ask big name corporations in the electronics and computer industry for the fund to maintain the specification documents and software implementations.
So we asked such companies to become a member of the would-be TRON Association.
Then president of Fujitsu, Mr. Takuma Yamamoto (he passed away in 2012) was very supportive of the idea of open architecture, and so he was asked to become the first chair of TRON Association, which he accepted.
The main electronic enterprises in Japan such as Toshiba, Hitachi, Fujitsu, Oki Electric Industry, NEC, Matsushita (today's Panasonic), and Mitsubishi Electric gathered and TRON Association as the not-for-profit corporation was established.
TRON Project became known worldwide, and in the early 2000s, corporations in the United States, Europe, and Asia began contacting us and told us that they wanted to contribute financially.
I have to explain a little bit about Japanese legal system back then. At the time, a non-profit organization needed to be either a Shadan Hojin or Zaidan Hojin (both are legal entities).
There was an issue regarding the tax, etc. to the contribution (or donation) even if the business was not-for-profit.
TRON Association was established as Shadan Hojin, but in the laws of the time, it had to be under control of a ministry of the government.
Because development and dissemination of the embedded system computer software were the stated goals, TRON Association became the Shadan Hojin of the Ministry of International Trade and Industry (MITI back then, now METI).
The ministry acts as a chaperon and checks if the not-for-profit organization was actually making money, etc.
However, this posed a problem for overseas contributors. They wanted to contribute to an independent non-profit organization. But contributing to TRON Association meant that they donated money to an organization under the control of a Japanese government ministry.
T-Engine Forum that was an independent nonprofit organization (NPO) was established in 2002 as TRON Project got widely known outside Japan.
ITRON specification standardized the set of API of real-time OS. However, the implementation was allowed to be CPU-dependent to make the best use of the power of the CPU of the time (weak or loose standardization).
However, because the performance of CPU had improved around the beginning of the 21st century, and the trade-off point of the performance vs. the developing efficiency shifted, the voice to hope for a stricter standardization of TRON RTOS family has grown for the middleware distribution.
However, the demand for ITRON specification OS running on legacy CPU remained.
So TRON RTOS family now has two members.
ITRON with the loose standardization approach remained with TRON Association.
A new RTOS member, T-Kernel was to be developed for more powerful CPU cores, and it would be based on a single source repository.
It was convenient to have both organizations running in parallel since T-Engine Forum could now take care of T-Kernel.
It was a pleasure to see overseas companies such as IBM and Microsoft, or national technical laboratory such as VTT of Finland join T-Engine Forum over the years. We continued an open architecture real-time OS development.
TRON Association and T-Engine Forum thus ran in parallel for several years. But the opinion grew among the supporters, that we should consolidate the activity. TRON Association was dissolved in 2010 and T-Engine Forum continued as the main supporter organization of TRON Project.
To begin with, the name "T-Engine Forum" is a name selected for a “forum to make CPU boards that run TRON RTOS" when there was TRON Association.
However, T-Engine Forum has continued the support of TRON Project alone for five years.
The activity to achieve the original goals of TRON Project such as Ubiquitous Computing or the Internet of Things (IoT) has become very strong, and the forum no longer is focused on the embedded OS and the development boards alone. So the reviewing the name "T-Engine" was a natural consequence.
Many T-Engine Forum members voiced the opinion that it would be better if we could use the original project name "TRON" in the forum's name.
Thus, the move and preparation to change the name to "TRON Forum" got in motion about half a year ago.
Luckily, the change got wide consensus from the members of T-Engine Forum.
As the chair of the forum, I declared that open source, open data, and open API will be three pillars of the forum activity in the coming IoT age at the TRON symposium last year.
So this name change happens at a good timing.
Let me summarize what TRON Forum does.
We do the following as before as the first pillar of the forum activity.
- Further development and maintenance of open source real-time OS kernel that has been maintained by the initial informal TRON Association, legally incorporated TRON Association, and T-Engine Forum.
- Activities such as management of source code, making better documentation, bug fixes, answering inquiries, holding seminars, the certification exam to test whether one understands the essentials of TRON RTOS will be continued as before.
- Holding training seminars and symposiums, attending various exhibitions such as TRONSHOW and others, and promoting TRON Project using the Internet.
Secondly, TRON RTOS family is an important part of the IoT, and TRON Forum clarifies its total goal is the IoT.
Final goal is ubiquitous computing and the IoT. The past activity of the forum is regarded as the preparation step of the important pieces for the final goal.
Open data is a new activity of the forum for the grand goal of the IoT.
The forum has participated in the activity such as Study Group of Open Data for Public Transportation.
To further the research in this direction,
we propose and codify the architecture in which the context-awareness to understand the real-world is done by the big data analysis of open data from many sources in conjunction with the large volume of sensor data collected and sent by many embedded nodes (where TRON RTOS runs).
The context-awareness is important in the IoT and ubiquitous computing.
It is necessary for successful context-awareness to combine the sensed data from many sensors for analysis after we understand the general structure of the world, how they work, etc. from the open data from many sources.
This is why open data plays an important role.
We shall establish a mechanism via which the data of public sector, including the dynamic data of railway operation, can be published.
We have done our parts in the Study Group of Open Data for Public Transportation. Also, we have been creating the barrier-free map that shows the routes on which physically-challenged people can move on using the open data of the Ministry of Land, Infrastructure and Transport and the local governments.
This has been done by TRON Enableware Study Group. We will integrate this work into the open data activity of the forum.
Thirdly, we shall begin the activity to call for the opening of the API of commercial devices that use TRON RTOS and other OS offerings. This is important for controlling devices in the IoT setting.
We shall create Open API Working Group in TRON Forum, and let it publish the API of devices created using TRON RTOS family and others. This shall help the deployment and adoption of the IoT.
Thus TRON Forum advances the activity of open source, open data, and open API aggressively as important ideas that support the IoT.
I am glad that I can report these new activity plans for the year starting in April.