In December 1999 just prior to the coming of 2000, the entire mass media in Japan was circulating reports without pause about Y2K, the so-called year 2000 problem that had not been much of a topic. It was dealt with in ordinary newspapers, of course, but also in weekly magazines and sports newspapers, and on television too--not just in news broadcasts, but even on programs aimed at housewives. I do not intend to get into a discussion of the Y2K problem here, but--leaving aside the outrageous assertions of people who said "when computers reach the year 2000, they will go out of order and stop, and great confusion threatening the human race will break out worldwide leading to panic"--it is in essence a specification error at the time of system design.
No matter what the various circumstances were (memory and disks were expensive, etc.), that the designers made the year into a two place number without considering that once software is introduced it will be used for a long time is the cause of everything. We could say this is a classic example of leaving something to be regretted for the future either because someone did not scrupulously decide something that would be important into the future, or because they made a final decision without deep consideration. For systems of the type that become society's infrastructure, systems design must be carried out on the basis of sufficient inquiry. Because it is a fact that once software is written it will be used for a longer period than one imagined, this has left us with the lesson that one must pay careful attention in deciding specifications even after considering them from various aspects of crisis management. Also, it has taught us that there is an enormous cost to be paid when you try to amend things later.
When we ponder whether there are not similar problems in addition to the specification for the year, I believe we could say that in Japan the problem of kanji [Chinese character] code also is exactly the same. Computers are entering into all sorts of places in our daily life. In particular, recently, in addition to the data processing and control fields that computers were applied to in the past, new applications such as the preservation of human knowledge have come into existence, and the period in which they will be incorporated in earnest is approaching. There are projects, such as the American Memory project at U.S. Library of Congress and our Digital Museum project at the University Museum of the University of Tokyo, that are attempting to preserve historical facts and academic materials.
In the case of the U.S., all that is needed are the 26 letters of the alphabet, but in a country like Japan with multiple character types, archiving knowledge requires that we include all the kanji that Japanese have created [and used] throughout history. In the TRON Project, from early on, we carried out a thorough investigation of the infrastructure to be used in Japan, and then last year we established a framework for 1.5 million characters, and inside that we completed a BTRON specification that can handle 130,000 characters. This was also put on sale in material form as the product "Cho Kanji."
It is clear that we will leave the roots of a calamity extending into the future if we lay down as is either JIS code, which has not secured a sufficient number of characters for kanji code, or furthermore Unicode, which is attempting to carry out character unification. It is not difficult to imagine to what extent revising already digitally archived items will lead to chaos similar to the Y2K problem. Both for the year and the character code, in spite of the fact that working with them one by one is simple, if we try making revisions to them afterward, we end up with a very tangled problem. I believe that TRON can greatly contribute in this field. I felt strongly through the Y2K problem that this must be asserted more forcefully.
If one asks why the TRON Project that encountered the year 2000 problem must do this sort of thing, it is because the goal our project is providing the base for the computer society of the future, and thus I would like to advance it further hereafter. It is truly regrettable that recently, using the clumsy Japanese-manufactured English guroobaru sutandaado, a mistaken proposal has come to the fore that says everything in Japan must be made to conform to international standards. All international standards are not necessarily proper. Just on the Y2K problem, that the year used with COBOL was a two place number and then ended up becoming an international standard has left great consequences well into future. In the 1960s, the problem had already become clear, [but nothing happened] even in spite of the fact that influence was exerted to try to make it a four place number. There is great resistance to revising a standard once it has been decided.
In particular, in the case of cultural and knowledge archives, there are many cases in which these are dependent on the regional language, and differences will appear in the way of doing things and in methodology. The appearance of differences itself is is an expression of the multifaceted nature of culture, which is interesting. In that sense, I cannot believe that the computers created with a focus on Europe and America are appropriate for archiving culture, and we are keenly aware we must strive even more for multilingual cultures centering around the Asian sphere. Thus it is our intention to create even greater developments through the TRON Project, which has successfully made it into the year 2000.
The above opinion piece by TRON Project Leader Ken Sakamura appeared on page 1 of Vol. 61 of TRONWARE. It was translated and loaded onto this page with the permission of Personal Media Corporation.
Copyright © 2000 Personal Media Corporation
Copyright © 2000 Sakamura Laboratory, University Museum, University of Tokyo